{"_id":"q-en-nixpkgs-2d32e27ece3b2564aa7527c9d0ede4c0cc9c646d3b77d997ebfa6469cb756e88","text":"Describe the bug Open indicator and see the following in journal The error at the top comes from now depending on settings schema (provided by ibus). This situation is annoyingly complex, because we cannot wrap . So what we try to do is create a wrapper with all the transitive dependencies of the plugins and call it day. We used to do patches to hardcode the gsettings, but this was far to much work to maintain and could be prone to developer error sometimes. I'm thinking the easiest solution would be to add all the transitive dependencies of to , and then wingpanel's can pick up all the proper dependencies. I think the only ideal request I could make to Pantheon upstream is to make use of SearchProviders from GNOME I believe this basically should be the very thing they want. They currently have a forked version of synapse's plugins system which IMHO appears to be inferior. To Reproduce Steps to reproduce the behavior: wingpanel-with-indicators the switchboard-plugin process get killed Expected behavior The switchboard-plugin process doesn't die. Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-8cdce653c42a871c4a6c41d51353d7b9b46f577216a03e4a0bb4ea8e4297091e","text":"Describe the bug Installing the latest SBT is not possible. To Reproduce Expected behavior Install is successful Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata"}
{"_id":"q-en-nixpkgs-f41ff943fdd0c5f258c68006e21f4cd0a4906dab0a41e4211f876c33595694af","text":"Describe the bug I'm trying to get Steam big picture to work with , xbox wireless adapter and a couple of xbox controllers. But if Steam is already running and I turn on my controllers, they successfully connect to the dongle, but steam doesn't detect them. If i first turn on the controllers, and then i start Steam, it recognizes them and they work fine. xow: I couldn't use the service file provided by the package because it would never start. I read somewhere that systemd is picky with symlinks, so i defined it myself, now it's starting fine. I also had to blacklist else xow would crash. Steam: To Reproduce Steps to reproduce the behavior: sure xow is running fine Steam on steam controllers to see if they work (they dont) Additional context I think xow is working fine. I tried the latest version (0.5) and the last commit, importing udev rules from the package and adding them to I also tried adding my user to the group. It doesn't matter if steam is running in big picture or not, i also tried steam native, still not working. Metadata Edit: Steam won't detect controllers even when using the wire. And also i tested the controllers with and and they work as expected. So i'm pretty sure the issue must be with steam.\nThis might be an upstream problem with , have you tried using it on a non NixOS system? I also maintain (a kernel driver for Xbox controllers) and I don't have this problem with Steam. Maybe try to test if there's still a problem on your system with xpadneo?\nYou might also want to try enabling xow using the NixOS module instead of writing a custom service:\nYes, i have another system with Arch and it's working fine there. Steam detects controller connections even when it's already running. That's what i tried at first, but xow wouldn't start, the service was hanging in an state. And I think it might be because when importing the service from the package, nix is creating a symlink to the service and it might cause some issues. Also I did some more tests with xow and controllers are working fine, , , this website https://gamepad- everywhere it works as expected, even Stardew Valley is able to detect them, but not Steam (well unless i restart it with controllers connected, then they are detected and working). It's really weird, I am able to pair them, xpadneo detects them, but they are totally unresponsive, everywhere. I uninstalled xow, and removed the wireless dongle when i tested this. Since xow it's working fine in browser with https://gamepad- and with some games, then udev rules should be fine , right? I would say it's an upstream issue with Steam, but it's working fine on Arch, so there might be some issues with some libs?\nThe xow NixOS module is fixed now. The problem was that NixOS specifically bypasses the Install section of service files and the module wasn't setting WantedBy for the service explicitly.\nI am wih something similar here, I am using the gamepad via USB , before the config all software detected the control, and I was able to play without problems except for steam, which prevented me from using the big picture, after enabling this option I have the same behavior as steam for all applications, if I plug in the control it does not work until the next reboot. Any suggestions on where to debug to understand why this behavior? Dmesg and lsusb look ok.\nHaving a similar issue, since I'm not using xow or anything but kernel drivers: Steam isn't able to detect a hotplugged Xbox one wired controller. It works perfectly fine if I plug it in before starting Steam. I can also confirm this behavior works on Arch.\nthat's my situation too\nI guess no news on this? :(\ni'll take a look, probably something to do with udev rules. i dont have a xbox controller though, only a steam controller\nThis problem sufficiently annoys me at this point that if there are any instructions I can follow to provide information I'm happy to do so. Thanks for offering to look. I am currently in possession of 2 xbox controllers, several controllers that show up as xbox controllers, and some controllers that don't show up as xbox controllers. I have this issue with all these controllers to the best of my recollection.\nthe steam udev rules are almost the same as on other distros, though arch does i'll try borrowing a xbox controller. does this happen with only the xbox one or does it happen with 360 too\nsadly i can't find a xbox one controller. the udev rules provided by the steam package are the same as on other distros, maybe it has something to do with our other udev rules. may also have something to do with bubblewrap\nIt seems to be related to linux usernamespaces which to my understanding remaps the id of the sender of udev events (from 0 to a big number) leading udev to ignore these messages if it's not zero. Running directly on my system and under (running as root and as a normal user) led me to believe this (). This issue may need to be fixed with bubblewrap, I don't think something can be done here. Unfortunately I don't have a solution :sweatsmile: PS : I don't know if fixing this issue will make everything work, .\nthese look related\nI merged someone who has this issue please look into using the patched systemd with the steam fhsenv\nI've just tested on latest NixOS unstable. A hotplugged Xbox One controller was successfully detected by Steam Big Picture. Thank you!\ni haven't done anything that would have fixed this\nWell that is exceptionally weird! Because it's definitely working for me, looks like an unrelated PR fixed it? This is straight out of my terminal, after starting Steam with no controller, disconnecting and connecting again:\nThis must be a change to how steam recognizes controllers. The program from still does not show events when started within There is still something odd: My joystick Logitech X52 HOTAS is recognized as a gamepad, while when I run steam packages on flatpak, it works properly. I don't know what they do differently in the flatpak version to get it working. Maybe using libgudev or a flatpak specific version of udev in the runtime packages? I have tested adding patched systemd to the fhs of steam, but this did not change anything (maybe some other udev library version was shadowing it)\nI marked this as stale due to inactivity. &rarr;\nI still have this issue, with both an xbox 360 controller with the wireless adapter as well with an 8Bitdo SN30+ in both wired and wireless mode. Starting steam with any of these already connected works, but if one gets connected after starting steam (or worse - shortly disconnects while playing) it will not be reconnected until a Steam restart. However, with a Steam Controller hotplugging works just fine! This happens with Steam Big Picture mode (and also in games, but I guess Big Picture mode is more readily reproducable). I assume I am doing something wrong as others have it fixed, is this really no longer an issue for others? Happy for any pointers and to assist in debugging. (my config is at in case it is needed, i am currently on )\nI assume I am doing something wrong as others have it fixed, is this really no longer an issue for others? I have the same issue, with PS5 DualSense controllers. I debugged a bit during the christmas holiday, based on some info in either this issue or some other... By applying a patch (linked from the issue) I was able to see udev starting to detect hotplugging the controller in the Steam FHS env (it didn't detect these events by default), but still Steam wouldn't pick up the hotplugged controller.\nI found my experimental config snippet from back then: AFAIR, I was worried that steam itself didn't use the patched libudev, but I'm pretty sure I looked up the process map and found it had (only) loaded the patched libudev, so I'm not sure what's going on. Didn't look at it since.\nIf anyone just wants to throw money at this problem I highly suggest the to solve it. The real win with this gamepad is that it always shows up as connected to your PC, even when it is powered off. It achieves this by coming with a usb dongle that shows up effectively as a wired joystick regardless if the gamepad is being used or not. Perfect solution for this problem... I have 3. I'll mention a potential con, though: Make sure you have good line of sight between the usb dongle and the gamepad (don't worry, it comes with a ~1m usb extension cable to allow good line of sight) and aren't too far away (3-4 m I would guess) or else the signal is really spotty. In my last setup I used this through the wall and it absolutely wouldn't work for any serious gaming, but now I have the pc in the same room as the gamepads with direct line of sight and it is perfect. But yeah... a real solution would be better and still appreciated by me. Just thought I'd post this in case anyone could benefit from it. :man_shrugging:\nAnother alternative is to use : in the config: then install and run it: I didn't have any issues since doing it since flatpak was updated to = 1.12\nI still have this issue on Nix's version of steam using nixos-unstable. Enabling xpadneo did not fix it either.\nmay have fixed this\nSeems like didn't fix it. I see that it is in nixos-unstable and it seems like Steam still displays the same issue with my DualSense controller. I need to restart Steam before it detects it\nI am having this issue on every controller I have: Steam Controller Nintendo Switch Pro Controller PS3 Controller Gulikit King Kong Pro 2 Controller in all 4 modes (Switch, XInput, DInput, Android/iOS) USB adapters for NES, SNES, and Gamecube controllers It occurs when using both Bluetooth and USB, so I am 90% certain it has nothing to do with which controller you are using. The behavior is the same for all of them; when I connect any USB/Bluetooth controller it is not recognized or noticed by Steam if it is already running, but only when Steam is starting up while the controller is currently plugged in. Here's some information on what happens when I use the Gulikit controller (XInput mode via bluetooth): Controller connected and Steam is not started At this point, and are now created Now I start Steam And reports: And now and are created (???) Disconnected the controller with Steam running Nothing new in I reconnect the controller with Steam running* And and is back TL;DR - So it looks like Steam will normally start up and find the controller and then it will create the virtual input devices and . However, it won't create these devices when they are connected after Steam is already running. The question I have is why this happens and what can be done to fix it. It looks to be a NixOS specific issue as the problem doesn't appear to be on other Linux distributions like Arch (just reading off of what's in this thread, I haven't tried myself). If anyone wants me to try anything and test it on a fair variety of USB/Bluetooth controllers please let me know. I think something might be wrong with how the FHS wrapper () seems to be binding the devices that prevents new devices from being seen. EDIT: I just confirmed the and device files are appearing and disappearing in the environment that Steam runs in, so the FHS is actually properly creating those devices. The only difference I can see is that they are owned by in the FHS environment (versus in the normal environment). I can still properly access both. I did see : However, that would imply this is not NixOS specific and instead is default Steam behavior. No indication of how to change that default behavior though, so I can't find a way to confirm it.\nThis seems related:\nJust as an datapoint: The Steam Controller is actually one that does not exhibit this issue for me, at least not when wirelessly connected with the Dongle that came with it. If I connect it via cable then it is the same as usual, only recognized when restarting steam. Also if connected via dongle and I replug the dongle, the controller only gets recognized when restarting steam. But if the dongle stays put and I shutdown the controller (by holding the Steam button for a few seconds) and start it again it works just fine with steam running. I guess it might be similar to the controller described in\nI was surprised as well, and a little bit of digging reveals that SDL2 being inside flatpak and disables the unreliable udev-based discovery. And disabling it manually ! I opened to fix this.\nCan confirm that it fixed the issue for me. Good job in discovering this\nThis is resolved for me now as well. Thank you all! :heart:"}
{"_id":"q-en-nixpkgs-0b744d30b40fbe12cc23d4b626b623090da0633469a308edeb5aceb85e1c7f10","text":"In a NixOS 20.09 guest on a VMWare Windows host, with dwm as my window manager, the mouse cursor neither moves nor registers clicks when . To Reproduce Steps to reproduce the behavior: NixOS 20.09 in VMWare Workstation Player 16 VM (from the minimal ISO) xserver and dwm and in as normal user with a minimal can you have a look?\nI can reproduce this. It does work if you login as but fails if you log into a normal user account. It also fails in .\nOh boy, even in the grown-up DM. That's bad.\nDid this fix the issue for you? ? I'm having issues my self. I also posted in reddit to see if maybe I didn't do something correctly.\ntry the PR I just linked. My mouse problems went away when I compiled the kernel with CONFIGPS2VMMOUSE.\nThank you it works! This will be great to have. This might be a silly question, but is open-vm-tools 11.3.0 required for this to work? I didn't have time to try 11.2.5 on the stable channel. For anyone reading this the name value is MOUSEPS2VMMOUSE, might save you one failed kernel build before you figure it out. Here's my nix config that I used to allow non-root user to use the mouse in Xorg running on vmware fushion on MacOS - This config was a \"get it to work\" attempt with no regard to what the ramifications of how I went about it. My next step is to figure out what it all really means and see if I went about it correctly."}
{"_id":"q-en-nixpkgs-e2337e9ea43ab378c1391ba0d87fe58330e4448b7b9b179648ed47ed7b827115","text":"Describe the bug from should have fledgling Ozone/wayland support. Users on other distros are having success. However, I am not. Looks like it might be a Nix-y thing: and no window opens.\nI had it working once: (didn't re-test lately tho) It might also be a hardware/driver thing (changing might also help if it's related to any local state). Do you get any additional output (I usually get way more output)? Or it might be due to one of the many upstream bugs (). But if this is indeed related to Nix investigating/fixing it would be very nice.\ntoo, saw it working once. , local state does affect it. I was having a totally different crash until I deleted . used to get more output, but no, that's everything, verbatim what I get from launching yesterday. It very well could just be upstream, but I'm getting jealous of adhoc reports of those builds working on other distros. Unfortunately I don't have a lot of time/ideas.\nOk, strange. I just tried this again with 87.0.4280.66 and got the following output (in my case this looks as with previous versions that crashed though). <details<summaryConsole output (the 2nd time no browser window appeared anymore)</summary<pre[nix-shell:~/google-chrome]$ which google-chrome-stable /nix/store/j3v003nw7mkhdzys4wih1zvw2lsvm8cq-google-chrome-87.0.4280.66/bin/google-chrome-stable [nix-shell:~/google-chrome]$ google-chrome-stable --enable-features=UseOzonePlatform --ozone-platform=wayland [(43)] (initialize): ANGLE Display::initialize error : Could not open the default X display. [(773)] EGL Driver message (Critical) eglInitialize: Could not open the default X display. [(1322)] eglInitialize OpenGL failed with error EGLNOTINITIALIZED, trying next display type [(43)] (initialize): ANGLE Display::initialize error : Could not open the default X display. [(773)] EGL Driver message (Critical) eglInitialize: Could not open the default X display. [(1322)] eglInitialize OpenGLES failed with error EGLNOTINITIALIZED [(20)] GLSurfaceEGL::InitializeOneOff failed. [(150)] Exiting GPU process due to errors during initialization [(621)] Failed to call method: org.freedesktop.DBus.Properties.Get: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(621)] Failed to call method: org.freedesktop.UPower.GetDisplayDevice: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(621)] Failed to call method: org.freedesktop.UPower.EnumerateDevices: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(374)] InitializeSandbox() called with multiple threads in process gpu-process. [nix-shell:~/google-chrome]$ google-chrome-stable --enable-features=UseOzonePlatform --ozone-platform=wayland [(621)] Failed to call method: org.freedesktop.DBus.Properties.Get: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(621)] Failed to call method: org.freedesktop.UPower.GetDisplayDevice: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(621)] Failed to call method: org.freedesktop.UPower.EnumerateDevices: objectpath= /org/freedesktop/UPower: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.UPower was not provided by any .service files [(43)] (initialize): ANGLE Display::initialize error : Could not open the default X display. [(773)] EGL Driver message (Critical) eglInitialize: Could not open the default X display. [(1322)] eglInitialize OpenGL failed with error EGLNOTINITIALIZED, trying next display type [(43)] (initialize): ANGLE Display::initialize error : Could not open the default X display. [(773)] EGL Driver message (Critical) eglInitialize: Could not open the default X display. [(1322)] eglInitialize OpenGLES failed with error EGLNOTINITIALIZED [(20)] GLSurfaceEGL::InitializeOneOff failed. [(150)] Exiting GPU process due to errors during initialization [(374)] InitializeSandbox() called with multiple threads in process gpu-process. [(122)] ContextResult::kTransientFailure: Failed to send GpuChannelMsgCreateCommandBuffer. </pre</detailsYou're right, I just tried the exact same version but with instead of and it works pretty flawlessly (videos don't play but the rest seems fine). So this issue only seems to affect Google Chrome. Also: With Chromium I don't get any of those error messages on the console that I get with Google Chrome (only the usual ones as in your initial post, not the ones related to ANGLE/EGL/X11).\nSince we will lose some Google account related stuff in chromium quite soon[1], I am looking into using Chrome for some stuff. At the moment I am using Chromium via and it is wonderfully fast and smooth (sway, nixos-unstable, intel-gpu), even though I have to disable something about video acceleration[2]. None of the three Chrome builds start with as reported in this Github issue. When also adding it runs (as a Wayland window), but the performance is bad. [1] ,-By%20Joey%20Sneddon&text=Users%20of%20the%20Chromium%20web,builds%20from%20March%2015%2C%. [2]\ncould you also report your Wayland compositor and GPU? Maybe that's related. I did test this on Sway and Weston (IIRC) with an old AMD Radeon HD 6950 GPU (Mesa's Radeon R600 Gallium3D Driver). I've also started following a few upstream issues that could be related: (seems Nvidia specific though) (reproducible on wlroots-based compositors) I'm still not sure if this is an upstream issue, GPU driver or Wayland compositor related issue, or Nixpkgs specific (either due to our Chromium packaging, due to some specific system dependencies, or due to issues with our Nix store paths). if you want you could also try to investigate this with e.g. (IIRC I briefly had a look but it's not a priority for me since I don't use Google Chrome).\nI am on sway/wlroots/intel-gpu. I'll look into strace, but I fear my Linux debugging skills are a bit rusty.\nI am not able interpret the output in any meaningful way, but I was able to get some more log output: conincides with\nWayland works on latest unstable chrome, but rendering falls back to software. Log Messages\nIt stopped showing window once I close and re-run chrome, but after removing it starts working again.\nCurrent output, just for the record (GPU driver: , flags: ). Current output of (only works via ): Note: This obviously shouldn't try to open the X display. Current output of (doesn't work at all):\nI have it working on stable 91.0.4472.114 with just the ozone platform flags, but there's no hardware acceleration at all.\nthat might be equivalent to the case then (which also works reliably but without hardware acceleration). For M92+ the situation is currently even worse as the GPU acceleration seems to be broken. I need to use to even launch Chromium or Google Chrome (unfortunately I lack time for a proper investigation but at least there are already some upstream bug reports).\nWith Google Chrome M92+ the Ozone/Wayland support seems to be completely broken on NixOS (coredump during startup): Chromium is still fine though. I've opened since there's also a nasty GPU acceleration regression now."}
{"_id":"q-en-nixpkgs-14c0ffae6d4286cf049b206ff6a9f2f065eb3603f0e5da6f2d8eba1f8c41aa1c","text":"Describe the bug\nThere is a version update available\nProposal to increase build timeout\nThat update PR includes a workaround (using older rust).\nKeeping this open until the channel progresses.\nupdated to .\nCould setting help?\nwe now have an actual failure on master cc\nRelated\nfixes the build of thunderbird.\nChannel advanced."}
{"_id":"q-en-nixpkgs-fd9bd2b6e8eee3e840944dd46c31b8c338c86998fdf9bcc6c5d34e71e0d61448","text":"Describe the bug version 1.18.0 fails to build. To Reproduce Steps to reproduce the behavior: Expected behavior Build properly completes. Logs Full build log: Additional context This is blocking rebuilds. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(anna): nixpkgs: Maintainer information:\nI'm experiencing this as well.\nmight be due to spansp upgrade in"}
{"_id":"q-en-nixpkgs-378f1ad9a8a2be357b8db3d3d9e21adb1434795a6708831de914e2e1f96347c7","text":"Describe the bug It appears that the binary is not included in the output of the derivation. To Reproduce Steps to reproduce the behavior: in NixOS 20.09. a video, e.g. by adding a solid color with “Open Other” → “Color” & dragging it to the timeline. to export with default settings. Expected behavior Video is exported successfully. Actual behavior Export fails with the following log message: Workaround separately. file after failed export. to create the exported video. Notify maintainers Metadata Maintainer information:\nThanks for your report, you are right, export is broken. But was also not included in $out/bin in older versions, so I guess there is a path to mlt missing somewhere, so shotcut does not find melt. We have to find out where.\nfound it, was changed to , just preparing a PR\nI have to apologize for not testing the export after updating to shotcut-20.10.31\nThanks, !"}
{"_id":"q-en-nixpkgs-50de2949aa3f53eb18d0829ddfd27ff661bcef97ef00043d4e325b2f61ef8549","text":"This issue tracks the conversion of from Docbook to CommonMark per . If you intend to work on this conversion, please comment below! here that you intend to work on it. (Please also post if you stop working on it.) a new branch of nixpkgs master. : change to . to . the contents from Docbook to . Follow . Use existing sections for inspiration. from the folder to build the manual. the manual on your computer at after running from folder: it looks good, submit a PR and reference this issue.\nHi Opened PR for this."}
{"_id":"q-en-nixpkgs-793705aec2024e9f515fdeaf5764a25ef01905be7b67abc3e0de07a5c390abf4","text":"This issue tracks the conversion of from Docbook to CommonMark per . If you intend to work on this conversion, please comment below! here that you intend to work on it. (Please also post if you stop working on it.) a new branch of nixpkgs master. : change to . to . the contents from Docbook to . Follow . Use existing sections for inspiration. from the folder to build the manual. the manual on your computer at after running from folder: it looks good, submit a PR and reference this issue.\nI will also work on this"}
{"_id":"q-en-nixpkgs-dc1aba34c22f500a90190bef5268cffb3f0199d5264e8a38366d7e282f91da3e","text":"Describe the bug Suspend to disk / hibernation in combination with ZFS may lead to filesystem corruption. To Reproduce The problem only shows up occasionally about once on every 30 startups. I had been using suspend to disk on my laptop without problem for several years. After changing from BTRFS to ZFS the import of the ZFS pool occasionally stalls when resuming from hibernate. A message says that the pool is corrupted and gets reconstructed. Once the pool was corrupted irreparably. Root cause ZFS does not support filesystem freeze and thaw under Linux at the moment. (see ) The problem is that when resuming from hibernation the initrd tries to import the pool (it also tries to fix a corrupted pool and effectively modifies the pool/partition). This happens before checks for a possible resume from hibernation and before the system returns to the state in which it had been suspended. Expected behavior It may be possible to fix the problem by reordering the sequence of actions in initrd. Unless there is a fix for this NixOS should disable the combination of ZFS and hibernation. (More precisely, initrd should either contain the setup scripts for ZFS or the scripts for return-from-hibernate). At least there should be a warning.\nBTRFS is native to Linux, so its suspension mechanism and Linux's were (presumably) always set up to work with each other. On the other hand, ZFS's suspension code has not yet been hooked up to Linux's properly: I agree that— One potential problem seems to be that there seem to be multiple packages that can cause suspension. In my NixOS configuration, I tell and not to suspend, but I don't know whether that's everything that might cause suspension.\nThanks for adding the link. I've a corresponding line of warning here:\nDo you suspend to disk with a zvol swap? Because I'm running with root on ZFS on my laptop and I have never encountered problems with suspend to disk, but I'm also using a separate regular swap partition.\nI for one use a regular swap partition. I recall hearing that swap-on-ZVOL is dangerous independently of hibernation.\nI marked this as stale due to inactivity. &rarr;\nWould it make sense to automatically add in cases where zfs is in use?\nThis. A note in wiki or manual will be good too. EDIT: Done ().\nI marked this as stale due to inactivity. &rarr;\nLet's see if we can close this with\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-fd2736ed7c10f5ec4ab3af3e6867151035a19c471c0fd7e4ebc983c5ab98c6bb","text":"Describe the bug is missing a menu entry to launch the application. To Reproduce Steps to reproduce the behavior: Expected behavior There should be a menu entry to launch the application. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata\nsee (probably shouldn't have been closed)"}
{"_id":"q-en-nixpkgs-4321184926d9f45603af558c6545dd1ab2bbd1efa8b5187f173cb9947b77efc5","text":"With the update to 1.48, rustc is missing some host platform libraries when built as a cross-compiler. Most notably, is missing which causes cross-compiled packages with scripts to fail with the following error: The native (host == target) compiler is not affected. The diff between the native and cross directory is as follows: This seems like an upstream regression. Interestingly, the list of missing libraries exactly matches the list of libraries installed in . cc\nlooks kind of suspicious.\nOr actually looks like it would explain the issue. I think we need to add the host platform to targets as well.\nThat makes sense to me and sounds like a good solution."}
{"_id":"q-en-nixpkgs-edf5f18f1eb27917c21f9e8f5c4f3e20463e766de3fdcd167327568dbb5b5c89","text":"Describe the bug on a NixOS configuration with systemd-boot as bootloader (the default for NixOS installs on UEFI systems) fails with this error: I checked that the file exists, so what's missing in the kernel since it reports \"No such device\"? Steps to reproduce This happens with nixos-20.09 and nixos-unstable, but not nixos-20.03. (Just change the command above.) Expected behavior The build succeeds, like it did before (on nixos-20.03). Screenshots If applicable, add screenshots to help explain your problem. Additional context Add any other context about the problem here. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata Please run and paste the result. Maintainer information:\nDisabling ~systemd-boot~ efi settings from results in So the insmod error is the same, but now the init script doesn't die on\nAh, the default value of is , and with that value it actually succeeds. So it's not really the systemd-boot part of the config that is broken.\nOh, this comment recommends NOT to set , because it can wear out the NVRAM:\nAh, the default value of is false, [...] Well, except that sets it to true, so it is enabled by default for NixOS EFI installs.\nInterestingly, the uses . That results in success!\nNow the question is, where to set (or )? I see at least these places where conditionally adding it could make sense: Call site: Option declaration, in qemu- Implementation, in qemu- In the systemd- module:"}
{"_id":"q-en-nixpkgs-52f0b29512d8113229bbca44beefd41713a84f961206e4bb8afbfa7097bf89fb","text":"current version: unstable-2017-10-06 desired version: 0.2.0 maintainers:\nAn option in the service definition to overwrite the package would be nice, too! :smile:\nWhy is this needed? -options are usually implemented if modifying the package in question with an overlay is a bad idea because it would mean that a lot of leaf-packages would require a rebuild as well. An example for that would be .\nAh, okay.\ncould you please update this package?\nOh sure, thank you for the ping :)"}
{"_id":"q-en-nixpkgs-f6af86c6086db1457f2ecba4f707b5d32d2c1f0a4c3cdca2bc804911448ec4b3","text":"Describe the bug Firefox crashed because my disk was full. I wanted to submit a crash report, but the site the documentation mentions don't exist. To Reproduce Steps to reproduce the behavior: in firefox Expected behavior firefox works like an official release Screenshots ! Additional context Add any other context about the problem here. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information:\nFirefox is compiled without crash reporting support by default: I don't know why, though.\nProbably because it could be considered tracking. But i think it's optional to actually send the crash reports and they don't contain personal data and are very useful to fix crashes. We would need clear guidelines which kind of tracking is OK.\nI don't have time to open a discussion about that. Since it crashed again, i tried with a workaround: rebuild fails with this error. i guess it's due to the structure of the firefox packaging. the option is in the while the package is defined in any idea?\nLooking at the definition quickly, I'd try\nThe build starts with that, but fails exactly at the crashreporter. I guess no one has tested that in a while. It took 33 minutes to get to that point with a i9 cpu (16x 4GHz). 32 GB RAM would be good, 16 will SWAP.\nI got it to build, on a second attempt Anyway, I don't think it will be that easy to actually use the reporter. For example, I don't think our default build keeps debugging symbols, and I don't expect it can really report anything useful without them.\nI marked this as stale due to inactivity. &rarr;\nI just built with crash-reporting enabled and everything seems to be working perfectly. Is there any reason we can't turn this on. In my case the reports are not even sending without asking first. If tracking is a concern may we can change the default from to .\nSince I don't hear any clear objections I have opened to hopefully push this issue along."}
{"_id":"q-en-nixpkgs-ec0156487de99f338561e4c78ac9341b083c856cb9d543fc1c63ca93d0f88e5c","text":"Describe the bug The Emacs packages for don't successfully load. To Reproduce Run a from the following : After opening Emacs from inside the and running , we get the following: Expected behavior The packages should be properly loaded, and they should offer syntax-highlighting for select languages. Notify maintainers No one has any particular responsibility for this package. But might be interested in this. Metadata Please run and paste the result.\nI'm currently using this, but it stills need some work. Languages grammar are downloaded instead of being built (each language should have it's own separate nix package), and I'm not happy about the but if the grammar bins are in directory will put them in the .\nI marked this as stale due to inactivity. &rarr;\nUnstale\nfacing the same issue\nSame. What's going on with this issue?\nSo, I've been looking into this a bit and kind of understand what should be done, but haven't been able to create enough code for a working PR yet. So, basically what you want is to build a rust package from here: The result of the build is and this needs to be built as a dependency for so the file is available for it as . I think it should be enough to link whatever file gets compiled from to the if I read the code correctly. Then, the next is to link all the compiled tree-sitter grammars for the emacs package to find, something copied from how the neovim plugin does this: I'm not completely sure where to build all these packages and I'm running out of my personal time to get this working together. What I see in the current state is many packages linking as a dependency, but is just building the emacs package, not using to build whatever is in the directory and linking the resulting library for the lisp libraries. And the grammars are already in the nixpkgs, you just need to link them to a place where the emacs tree-sitter plugin can find them.\nHey all. I'm still working on fixing this due to me needing tree-sitter for my emacs setup and wanting to understand nix more. Here's the work so far: It now compiles the rust module correctly, and uses the linking strategy from for the languages. It still tries to download files from github, so I either don't link the files to the correct location, or I should patch a method to prevent the download process. I think the best solution would be to use the grammars already available in the monorepo the same way as does. My first try with them was with , but some of them won't work together with the emacs library due to licensing issues. And I also got myself to the download problem, but I think this is not due to files not being the pre-compiled ones from github, but again about some linking issues. If somebody wants to try this out and give me feedback, I'm all ears. I will probably spend some Christmas holiday spare time to make the PR happen. It is very close already.\nI made a PR and now it works at least on my NixOS laptop. It would still be nice if we could either utilize the grammars already available in nix, or at least we'd be able to compile them in the derivation instead of just downloading the binaries. I'd be happy if you could try this out and see if it works for you."}
{"_id":"q-en-nixpkgs-e9fa2576566659daa70cbbda1b4c3a44b729e153a0c857b5fc41f34ce7775fec","text":"Describe the bug gpgme fails to build on macOS To Reproduce <details<summary<codenix-build . -A gpgme</code(<iexpand to see the output</i)</summary<prewarning: Nix search path entry '/nix/var/nix/profiles/per-user/root/channels' does not exist, ignoring this derivation will be built: /nix/store/zwf2jgy24qs7b69n1rb5pg9ywwmi82f0-gpgme- building '/nix/store/zwf2jgy24qs7b69n1rb5pg9ywwmi82f0-gpgme-'... unpacking sources unpacking source archive /nix/store/sq6ypkha64hcskq8qqw8rv40d9k4y94f-gpgme- source root is gpgme-1.15.0 setting SOURCEDATEEPOCH to timestamp of file gpgme- patching sources applying patch /nix/store/d5svj9vzmvfvlfh67chy0yliwjk50wb5- patching file src/engine-gpg.c Hunk succeeded at 2989 (offset 28 lines). applying patch /nix/store/v28z94l2z3669df0vzwgsi4y53rvv5rc-0001-Fix-python-tests-on-non- patching file patching file applying patch /nix/store/k3dx8r5bp3pqd4dzrydj4yywg6c9p6ja-disable-python- patching file autoreconfPhase autoreconf: export WARNINGS= autoreconf: Entering directory '.' autoreconf: not using Gettext autoreconf: running: aclocal --force -I m4 autoreconf: tracing autoreconf: running: libtoolize --copy --force libtoolize: putting auxiliary files in ACCONFIGAUXDIR, 'build-aux'. libtoolize: copying file 'build-' libtoolize: putting macros in ACCONFIGMACRODIRS, 'm4'. libtoolize: copying file 'm4/libtool.m4' libtoolize: copying file 'm4/ltoptions.m4' libtoolize: copying file 'm4/ltsugar.m4' libtoolize: copying file 'm4/ltversion.m4' libtoolize: copying file 'm4/lt~obsolete.m4' autoreconf: not using Intltool autoreconf: not using Gtkdoc autoreconf: running: aclocal --force -I m4 autoreconf: running: /nix/store/ymq2bmizcmzffq8qwi11lpm2sfl0d9r2-autoconf-2.70/bin/autoconf --force warning: The macro ACCONFIGHEADER' is obsolete. You should run autoupdate. ./lib/autoconf/status.m4:719: ACCONFIGHEADER is expanded from... the top level warning: The macroACGNUSOURCE' is obsolete. You should run autoupdate. ./lib/autoconf/specific.m4:312: ACGNUSOURCE is expanded from... the top level warning: The macro ACTRYRUN' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:2996: ACTRYRUN is expanded from... m4/glib-2.0.m4:8: AMPATHGLIB20 is expanded from... the top level warning: The macroACTRYLINK' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:2919: ACTRYLINK is expanded from... m4/glib-2.0.m4:8: AMPATHGLIB20 is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... m4/gpg-error.m4:27: AMPATHGPGERROR is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... m4/libassuan.m4:17: AMPATHLIBASSUANCOMMON is expanded from... m4/libassuan.m4:158: AMPATHLIBASSUAN is expanded from... the top level warning: The macro ACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level warning: The macroACHELPSTRING' is obsolete. You should run autoupdate. ./lib/autoconf/general.m4:203: ACHELPSTRING is expanded from... the top level autoreconf: running: /nix/store/ymq2bmizcmzffq8qwi11lpm2sfl0d9r2-autoconf-2.70/bin/autoheader --force autoreconf: running: automake --add-missing --copy --force-missing installing 'build-aux/compile' installing 'build-aux/missing' installing 'build-aux/depcomp' autoreconf: 'build-' is updated autoreconf: 'build-' is updated autoreconf: Leaving directory '.' configuring fixing libtool script ./build- configure flags: --disable-static --disable-dependency-tracking --prefix=/nix/store/9357hqrwqm6wnvl6qnp1hw2h045nw2rf-gpgme-1.15.0 --bindir=/nix/store/r9355cgd5y62qh03h68b4a6lj3w3kcap-gpgme-1.15.0-dev/bin --sbindir=/nix/store/r9355cgd5y62qh03h68b4a6lj3w3kcap-gpgme-1.15.0-dev/sbin --includedir=/nix/store/r9355cgd5y62qh03h68b4a6lj3w3kcap-gpgme-1.15.0-dev/include --oldincludedir=/nix/store/r9355cgd5y62qh03h68b4a6lj3w3kcap-gpgme-1.15.0-dev/include --mandir=/nix/store/r9355cgd5y62qh03h68b4a6lj3w3kcap-gpgme-1.15.0-dev/share/man --infodir=/nix/store/a7im3pbpc9aa0v37wlclqnas2j5cx5kv-gpgme-1.15.0-info/share/info --docdir=/nix/store/9357hqrwqm6wnvl6qnp1hw2h045nw2rf-gpgme-1.15.0/share/doc/gpgme --libdir=/nix/store/9357hqrwqm6wnvl6qnp1hw2h045nw2rf-gpgme-1.15.0/lib --libexecdir=/nix/store/9357hqrwqm6wnvl6qnp1hw2h045nw2rf-gpgme-1.15.0/libexec --localedir=/nix/store/9357hqrwqm6wnvl6qnp1hw2h045nw2rf-gpgme-1.15.0/share/locale --enable-fixed-path=/nix/store/x5v8m4xf9sf07675db7cfi19g9qjp4y1-gnupg-2.2.24/bin --with-libgpg-error-prefix=/nix/store/4sdr1n9fvhpry7410jfyd87sdsvh5422-libgpg-error-1.38-dev --with-libassuan-prefix=/nix/store/c0pfl1d4dw61i3mwzc52akb9vbx223vf-libassuan-2.5.4-dev --disable-gpg-test checking for a BSD-compatible install... /nix/store/5m4875pw1zcd588n7a6j88vwsilvnm7j-coreutils-8.32/bin/install -c checking whether build environment is sane... yes checking for a race-free mkdir -p... /nix/store/5m4875pw1zcd588n7a6j88vwsilvnm7j-coreutils-8.32/bin/mkdir -p checking for gawk... gawk checking whether make sets $(MAKE)... yes checking whether make supports nested variables... yes checking whether to enable maintainer-specific portions of Makefiles... no checking build system type... x8664-apple-darwin20.2.0 checking host system type... x8664-apple-darwin20.2.0 checking whether make supports nested variables... (cached) yes checking whether make supports the include directive... yes (GNU style) checking for gcc... clang checking whether the C compiler works... yes checking for C compiler default output file name... a.out checking for suffix of executables... checking whether we are cross compiling... no checking for suffix of object files... o checking whether the compiler supports GNU C... yes checking whether clang accepts -g... yes checking for clang option to enable C11 features... none needed checking whether clang understands -c and -o together... yes checking dependency style of clang... none checking for stdio.h... yes checking for stdlib.h... yes checking for string.h... yes checking for inttypes.h... yes checking for stdint.h... yes checking for strings.h... yes checking for sys/stat.h... yes checking for sys/types.h... yes checking for unistd.h... yes checking for wchar.h... yes checking for minix/config.h... no checking for locale.h... yes checking for sys/select.h... yes checking for sys/uio.h... yes checking for argp.h... no checking for sys/time.h... yes checking whether it is safe to define EXTENSIONS... yes checking whether XOPENSOURCE should be defined... no checking for gcc... (cached) clang checking whether the compiler supports GNU C... (cached) yes checking whether clang accepts -g... (cached) yes checking for clang option to enable C11 features... (cached) none needed checking whether clang understands -c and -o together... (cached) yes checking dependency style of clang... (cached) none checking how to run the C preprocessor... clang -E checking whether the compiler supports GNU C++... yes checking whether clang++ accepts -g... yes checking for clang++ option to enable C++11 features... none needed checking dependency style of clang++... none checking for gitlog-to-changelog... no checking for cc for build... clang checking how to print strings... printf checking for a sed that does not truncate output... /nix/store/n0zyx5b5s1grq0jrcy2xvf7apvh25wry-gnused-4.8/bin/sed checking for grep that handles long lines and -e... /nix/store/pmbv3czhsa816msq51ngb2db61crjqw0-gnugrep-3.6/bin/grep checking for egrep... /nix/store/pmbv3czhsa816msq51ngb2db61crjqw0-gnugrep-3.6/bin/grep -E checking for fgrep... /nix/store/pmbv3czhsa816msq51ngb2db61crjqw0-gnugrep-3.6/bin/grep -F checking for ld used by clang... ld checking if the linker (ld) is GNU ld... no checking for BSD- or MS-compatible name lister (nm)... nm checking the name lister (nm) interface... BSD nm checking whether ln -s works... yes checking the maximum length of command line arguments... checking how to convert x8664-apple-darwin20.2.0 file names to x8664-apple-darwin20.2.0 format... funcconvertfilenoop checking how to convert x8664-apple-darwin20.2.0 file names to toolchain format... funcconvertfilenoop checking for ld option to reload object files... -r checking for objdump... no checking how to recognize dependent libraries... passall checking for dlltool... no checking how to associate runtime and link libraries... printf %sn checking for archiver support... no checking for strip... strip checking for ranlib... ranlib checking command to parse nm output from clang object... ok checking for sysroot... no checking for a working dd... /nix/store/5m4875pw1zcd588n7a6j88vwsilvnm7j-coreutils-8.32/bin/dd checking how to truncate binary pipes... /nix/store/5m4875pw1zcd588n7a6j88vwsilvnm7j-coreutils-8.32/bin/dd bs=4096 count=1 checking for mt... no checking if : is a manifest tool... no checking for dsymutil... dsymutil checking for nmedit... no checking for lipo... lipo checking for otool... otool checking for otool64... no checking for -singlemodule linker flag... yes checking for -exportedsymbolslist linker flag... yes checking for -forceload linker flag... yes checking for dlfcn.h... yes checking for objdir... .libs checking if clang supports -fno-rtti -fno-exceptions... yes checking for clang option to produce PIC... -fno-common -DPIC checking if clang PIC flag -fno-common -DPIC works... yes checking if clang static flag -static works... no checking if clang supports -c -o file.o... yes checking if clang supports -c -o file.o... (cached) yes checking whether the clang linker (ld) supports shared libraries... yes checking dynamic linker characteristics... darwin20.2.0 dyld checking how to hardcode library paths into programs... immediate checking whether stripping libraries is possible... yes checking if libtool supports shared libraries... yes checking whether to build shared libraries... yes checking whether to build static libraries... no checking how to run the C++ preprocessor... clang++ -E checking for ld used by clang++... ld checking if the linker (ld) is GNU ld... no checking whether the clang++ linker (ld) supports shared libraries... yes checking for clang++ option to produce PIC... -fno-common -DPIC checking if clang++ PIC flag -fno-common -DPIC works... yes checking if clang++ static flag -static works... no checking if clang++ supports -c -o file.o... yes checking if clang++ supports -c -o file.o... (cached) yes checking whether the clang++ linker (ld) supports shared libraries... yes checking dynamic linker characteristics... darwin20.2.0 dyld checking how to hardcode library paths into programs... immediate checking for windres... no checking whether clang++ supports C++11 features by default... yes checking pkg-config is at least version 0.9.0... yes checking for GPGMEQT... no checking for GPGMEQTTEST... no Package Qt5Core was not found in the pkg-config search path. Perhaps you should add the directory containing ' to the PKGCONFIGPATH environment variable No package 'Qt5Core' found configure: WARNING: Qt5 (Qt5Core) not found Qt Binding will be disabled. checking for swig... no checking for swig2.0... no checking for inline... inline checking size of unsigned int... 4 checking for special C compiler options needed for large files... no checking for FILEOFFSETBITS value needed for large files... no checking for offt... yes checking for uintptrt... yes checking if gcc supports -Wno-missing-field-initializers... yes checking for _thread... yes configure: checking for libraries checking for LARGEFILE_SOURCE value needed for large files... no checking how clang reports undeclared, standard C functions... configure: error: in/private/tmp/nix-build-gpgme-': configure: error: cannot detect from compiler exit status or warnings See `' for more details error: --- Error --- nix-build builder for '/nix/store/zwf2jgy24qs7b69n1rb5pg9ywwmi82f0-gpgme-' failed with exit code 1 </pre</detailsExpected behavior To build successfully. Additional context I'm working on my flake based config with nix, home-manager & nix-darwin. I updated my inputs today & after that, I couldn't build anymore. I cloned this repo & ran which failed with the same error from I also tried this workaround but it didn't help. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata*"}
{"_id":"q-en-nixpkgs-36d0b25cf4fb501da6dbbd5dae1672914984da477799320b737d414210cecef7","text":"At: This should also add as per Gitea's docs. ( ) Without this tag sqlite will frequently report that the database is locked and cause multiple internal server errors. See\nIf someone needs a quick fix in `:"}
{"_id":"q-en-nixpkgs-e0d1101fc1c370ba3413f449bd59b604c0cdf6eec82990828570083146da8dd0","text":", CVSSv3=7.5 (nixos-20.09, nixos-unstable) CVSSv3=7.5 (nixos-20.09, nixos-unstable) Scanned versions: nixos-20.09: ; nixos-unstable: . Cc"}
{"_id":"q-en-nixpkgs-46d046f221f9ab1d009937f952032b2875d5db56c4c7fe622413847e29ff5a3d","text":", CVSSv3=9.8 (nixos-20.09, nixos-unstable) Scanned versions: nixos-20.09: ; nixos-unstable: . Cc"}
{"_id":"q-en-nixpkgs-fa04d160eb922650121809ade9f0e6f288ba865516ccf8d9061aa8750c1e1985","text":"For this patch I to CudaText the global variable which you can change to False. to disable \"Help / Check for updates\".\nOK, I will use it with next release."}
{"_id":"q-en-nixpkgs-e75d5720c1e4d8172f7491568d7db6fbe922da70b362c789c1dd4d05474780a4","text":"Describe the bug sysklogd fails to build when built against musl-libc To Reproduce Steps to reproduce the behavior: Expected behavior Package should build. Notify maintainers no maintainers, because apparently this is ancient. Metadata Maintainer information:"}
{"_id":"q-en-nixpkgs-0cd534f1c8b9356205b5cd7e6cc55e45ff8c343a878ae7839499dbc8d726aa8f","text":"I tried adding to , then running and noticed in the diff this still uses , and non-string URLs. To Reproduce Steps to reproduce the behavior: Expected behavior The diff shouldn't contain non-string URLs, and not use via . Notify maintainers\nluarocks-nix needs to be updated. Its current state is messy (not clear which branch to use), I will update it."}
{"_id":"q-en-nixpkgs-5c6cb5f70052feaeffde83bb14fe695202121d8e3c56f9f6eedf983d27782d3a","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on github to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 1.5.2 desired version: 1.8.3 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR.\nI managed to use this overlay for now:\ncould you please take for a test run? Thanks\nChecked out the branch and it seems to be working :smile:"}
{"_id":"q-en-nixpkgs-e03a10c91c0ee54a8d9c2e6f75c8487264cc44287de61fe6ab217f558c121a35","text":"Hi, I'm testing nextcloud it seem there is an issue. I'm tying to install a fresh nextcloud and with the minimal example of the man : I get : janv. 29 16:56:31 cloud systemd[1]: Starting nextcloud-setup.service... janv. 29 16:56:31 cloud nextcloud-setup-start[]: Nextcloud is not installed - only a limited number of commands are available janv. 29 16:56:31 cloud nextcloud-setup-start[]: janv. 29 16:56:31 cloud nextcloud-setup-start[]: Command \"upgrade\" is not defined. nextcloud-setup-start try to update a new install same result with : = [ \"nextcloud-19.0.6\" ]; or system: `\"x86_64-linux\" host os: multi-user?: sandbox: version: channels(root): nixpkgs: I may not be calling all the maintainers in case it's a false alarm :-)\nI just ran into this too, with as my package. Running doesn't show an command.\nI ran the command manually, and now shows up. Not sure why it didn't install through nix, though.\nThanks, it works for me too. It must certainly come from the test. I'll check that later.\nCannot reproduce this on latest master and release-20.09. Can you please give me a minmal VM config that I can use to reproduce?\nSame config works in a test VM. I'll try to take a closer look next week if it's still relevant.\nI am running into this error as well with about the same config as . Specifically I'm using configuration without the nginx stuff (e.g. lines 5 to 27 commented out and not set) for on both nixpkgs master and release-20.09. I'll try it in a VM and report back..\nIt doesn't look like the builds of nextcloud and postgres are cleaned up completely, which is where it looks like this problem is coming from. Specifically, I had built nextcloud and postgres several times while I worked out the configuration settings I wanted. It turns out there were nextcloud and postgres folders in that still remained after I uninstalled both apps. Removing those two folders, then reinstalling these programs via nix fixed the error for me.\nWhen I've first set up Nextcloud i encountered the same issues is describing, but since then everything has run smoothly for me.\nIt might sense to verify the result of after the block, before executing , e.g. using and , where causes 's returncode to represent the boolean value of . The of is unclean. Instead of returning simply JSON, it also returns an informative message on the same FD: Although this should be fixed upstream in the long term (see also ), it needs to be stripped away for now (e.g. using ).\nmay prevent corrupted installations in some cases. But I'm still missing a minimal VM setup to reproduce.\nI thought I gave one. I 'm very new with nixos but I would be delighted to learn more. maybe you can give me a link or guide me to do this? or pass the assignees to someone that can reproduce it. thanks\nWell, that's the test VM's config that I already checked. Was there any existing state in /var/lib/nextcloud during your install attempt?\nI don't know I need to reproduce this on the machine in question. i can try\nhave you checked whether repeated calls are actually safe? Especially if you want to \"work out\" your options iteratively. Yeah please do that. You can also answer my question \"Was there any existing state in /var/lib/nextcloud during your install attempt?\". This is IIRC the main cause for problems like this.\nI don't have access to the machine but I've just started a vps and the same config works with the same steps. the problem would not have been solved :-) ? I'm going to look at the merge for the fun\ncool merge :-) I've used adminpass = \"\" on vps may be this is the trick. And ok I just read other posts.\ndo you really need me to reproduce the problem ? it's really very easy and I know you are active in others post who explain the problem. If you really need that I have a vps whith the bug.\nWell, I'm trying to check if there are any other causes apart from an existing state. Hence my question whether there was any existing state when you encountered the issue.\nI don't know in any case, /var/lib/nextcloud/* seems to be the same between the 2 generations\nHit this too: I'd misconfigured the postgres database (the nextcloud module doesn't setup a nextcloud user and role on the db, which I'm used to for other modules), so the service errored out with a database \"role doesn't exist\" issue. Once I fixed this, I hit this problem, which went away with a good old .\nTo me it seems as if most of the issues here are because the initial install failed and then the app wasn't properly installed, correct? Rather than letting this stall here even more, I'd write a few more notes in the documentation."}
{"_id":"q-en-nixpkgs-08f44b391fec0e3d4631b527d9b40f8de0825cb8650353feeb92515c6ded6840","text":"Project description Video DownloadHelper is a \"Browser extension to download videos from the Web\" Metadata homepage URL: source URL: license: gpl2 platforms: linux, darwin, (windows) other build scripts\nI marked this as stale due to inactivity. &rarr;\nFYI, I have been trying to figure this one out from source, but as isolates the build (which means it does not allow contacting any URL to download something), and one of the dependencies totally requires going to GitHub, I'm not able to build it. I am still on it, and if building from source doesn't work, I hope the tar option works as a solution (really want to avoid the deb option).\ni would recommend or to build the nodemodules from the package- lockfile. in this case, package- is missing, and must be generated with (the lockfile should be to the vdhcoapp repo ...) at least npmlock2nix allows to symlink extra files into nodemodules\nFYI, I have been trying to use the methods you mentioned, and I kind of figured out how to solve that GitHub dependency issue mentioned before (had to download it through ), but now I got into an issue where I don't get any specifics of what to do, unfortunately:\nI have been also trying to do it with the tar and the deb files. The tar requires an extra install step (be it for system or user), and I get the following error: For the deb, that does not require the extra steps, I can install it fine, but Firefox is not recognizing it. !\nin you would do something like to create a symlink from to (i found with ) if youre stuck, maybe share your draft. this cant be too hard to build from source\nJust in case, I've found this one which is way more advanced in building it. It even uses a patch to avoid using pkg (which was the issue I was having earlier). Plus, it is kind of similar to the way it is built on the AUR. The only thing is that I'm not able to build it as it is (copying and pasting it into ), so I need to see the code first and then check what can be adapted. I'd rather use this as a guide. I had already overwritten the I was using from source with the deb one.\nnice! thing is, that one was made with npm2nix which generates , which should be replaced with package-, stored in the upstream repo (vdhcoapp), to keep nixpkgs small the 5KB patch can be compressed to not tested, but im feeling lucky ; )\nIn fact that's something neat to adapt, going to try it :) In the functional part, it seems that it is not enough to just install it, you also need to apply the part, which at least is recognized by the extension, but provides an error with no description of what to do or what caused it: ! EDIT: Nevermind, it works. Weirdly, it took a while, but now it says it recognized it. ! So the preferred situation would be just installing this and making it work as expected, instead of running that extra install step. And this is where I am not sure how NixOS masks the /usr and /etc directories where the are located.\nshould work with and or but i always get which is a known WONTFIX\nFYI, I set this up on .\nThank you for packaging this into NUR. After installing from there I still had to do Otherwise neither chromium nor FF couldn't find it even after reboot.\nSure thing I wish I could find a way to indicate that when installing, but for now that's the only way it works. Now with 21.11, it should even work with Chrome (that specific browser was looking for a specific GLibC version packaged on 21.11).\ncan the NUR package be used on pure nixos given that the manifest should live under ? After installation and I still can't trick firefox into recognizing the plugin. EDIT: it's entirely possible that it's due to the fact that I install firefox via , since I can use the plugin from chromium\nYes, if Chromium is detecting the manifests after applying the user setup, then this might be something related to your Firefox setup. I will still try your recommendation.\nI couldn't make it work on Firefox unfortunately (NixOS unstable with NUR package and post-install cmd)\nLet's move the discussion to , as this is more related to the packaging issue. I will throw more details there.\nTo wrap it up - silly me forgot that I need to rerun the post install command every time (browsers?) update. Now works both on Chromium and Firefox.\nposted a way for Firefox through Home-Manager to set the VDHCoApp manifests through the system installation. . With that, I'm seriously considering sending a PR to Nixpkgs, just need to check and format the code properly.\ndid you find time to do further work on this? you have mentioned here, but also on the issue on your nur repo that you are considering to do a PR for the app to the nixpkgs.\nI have indeed, although it was more on re-running the process of creating the whole nix setup from scratch and use NodeJS 16. Now, my only concern (besides the code and many of my PR's not being reviewed) is that I don't want to keep maintaining more stuff on nixpkgs. If anyone wants to throw it, just go for it and I can help you, but personally I don't want to keep maintaining more packages because of lack of time.\nFYI, a new major version will be released (2.0) and it is a refactor on the building process, which might make it easier to install compared to the way it currently works.\nI curious how's the work on packaging version 2.x is going? Is there still intention to merge it into nixpkgs?\nI have been trying it sporadically. The wall I'm facing is regarding the . If I skip it, I don't see anything useful to add to the output, but if I try to bundle it, I get a issue. I was able to figure out the rest though: Another option would be finding a way to replace the usage of with something else to build the file that is generated during the bundling process.\nHi, i'm kind of a nixos rookie so sorry for asking, but: i tried installing the coapp first via but firefox did not recognize it (loads forever), so installed it via NUR, but now the browser wants me to update the coapp (which is pinned to 1.6.3 here it seems) so: how can i update this? ( i tried cloning it but running fails as node2nix runs into )\ncurrently there is no way unless you can get 2.X packaged. As mentioned on the comment above, I'm struggling with the bundling process that uses to get the new binary.\nok, thanks for the very fast reply! hmm thats unfortunate im rather busy the next week(s) so i can only keep my fingers crossed and hope you'll find a way in the unlikely event that i figure out a way after that i'll keep you posted thanks for the efforts Edit: but in case i find time, will your current efforts be publicly available somewhere? (not just the , but a more complete repo e.g. also containing )\nSure, you can find it .\nI was trying to get your 1.6.3 version working with 2.0.x, but I'm very new to packaging in nix and didn't get very far. So, I'm glad to hear that you are working on it. Good luck and let me know if I can help test in any way.\nHi, If anyone can add release 2.1.9 Best Regards\n, there is a reason version 2.X still hasn't been yet.\nGood news everyone, someone figured a way to package it and it's working, so I just adapted it to work the nixpkgs way. If you are interested in pushing it to nixpkgs (I really don't want to maintain it), you will need to pack (could be internal to the same vdhcoapp package) and .\nIm not experienced enough and/or attached enough to NixOS yet to take that on, but i would really appreciate if someone makes it happen. If i get into NixOS more in the future, i would also be willing to take it over at some point if needed\nwell done! Im not very experienced either, but i use nixos as my main system so at least im attached enough ;) i will try pushing it to nixpgs (as my first contribution) today"}
{"_id":"q-en-nixpkgs-1187b9de99884944b3b0c9021aa6490c2340fa7483a4f9ddca905f1e150a47df","text":"Describe the bug fails to build on master and unstable Notify maintainers\ncould you please take a look at ?\nThanks"}
{"_id":"q-en-nixpkgs-33fc5b8bd9c060fbcb345ce38042cddc1a6b99e4ce56cfe4e25a3943047682de","text":"Describe the bug I'm adding gimp like this in 20.09: However the fourier plugin is not being recognized. It should add: Filters/Generic/Forward FFT Filters/Generic/Inverse FFT But these filters are missing. Expected behavior FFT plugin should be recognized by gimp. Screenshots If applicable, add screenshots to help explain your problem. Additional context Add any other context about the problem here. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata Please run and paste the result. Maintainer information:\nI'm not sure about these specific plugins, but gimp-with-plugins comes with G'MIC, which includes a FFT filter under Filters -G'MIC Qt... -Available Filters -Frequency -Fourier transform"}
{"_id":"q-en-nixpkgs-e1779595121343e66c44df1782fdc5c7be2053ed46139b687c7c05db9a28c9f2","text":"Describe the bug Failed to build . To Reproduce in -rebuild switch Expected behavior Build successful. Screenshots  Additional context Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nNew warnings/error are introduced with new GCC releases.\ncould you please give a try?\nYes, it works."}
{"_id":"q-en-nixpkgs-0f5ead2b70f325b2203d08a8c000a7f84e33e75452a376bc3892a551ec010f67","text":"Describe the bug I have a nextcloud installation in the vhost root: which would redirect to for the NextCloud login. Also I enabled in the nginx settings that a SSL certifcate with the acme service should be used. However, no certificate can be issued - acme claims that it gets 404. To Reproduce Steps to reproduce the behavior: nextcloud and acme configuration to your configuration nix. E.g. will generate a nginx config that contains: first part of that configuration snippet also matches for the folder and hence a 404 is returned. Expected behavior Having SSL cert issued by Let's Encrypt. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information:\nOpened with a fix.\nThanks for the quick fix :)\ncan you confirm that this fixes the issue for you? If so, I'd merge+backport. Just in case you didn't know - you can do\nI did manually edit the nginx config in the nix store: Then run the acme command as standalone: So it works."}
{"_id":"q-en-nixpkgs-7ef3abb1390a145dc1caa7687c3c11ce882c671b9bb6e9b3d55b8a3204d7c959","text":", CVSSv3=7.5 (nixos-20.09, nixos-unstable) Scanned versions: nixos-20.09: ; nixos-unstable: . Cc"}
{"_id":"q-en-nixpkgs-1303a3970c0f5e7a094e20219021815488ba6112132cac38e58480860f7a34c2","text":"Describe the bug AdGuardHome has the meta platform , but is a . To Reproduce Try to run on (or any architecture other than ). It will fail with : Expected behavior should execute, so that it can be used. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information:\nI will try to get it instead of pulling a pre-compiled binary, but I'm not sure whether I'll get this solved. EDIT: Okay, that looks to be above my packaging skills, as the project consists of both a node component and a golang component, so it's nontrivial to package. An intermediate solution might be to use the architecture to determine the download URL, but that would make the src (and thus the hash) depend on the architecture. The downloads for different architectures .\nHotfix for aarch64 by using :\nI marked this as stale due to inactivity. &rarr;\nI also have the problem from the thread starter and I was wondering whether there is now official support for aarch64 (trying adguardhome on raspberrypi 3) or do I need to change smth in my config. I just use adguardhome as the inside of the systemPackages and I enable it. I also see this PR: .\nthe pr wasn't backported to 21.05. the pr will be in the 21.11 release which will be released today or tommorow afaik\ngreat, thx for the info :)"}
{"_id":"q-en-nixpkgs-bf100b66019276b2afc5e10eee1a928cae7941a064b00d110f6e11eaa9863deb","text":"Describe the bug As of this commit the package no longer allows overriding the JRE without failing to build. It still worked as of this commit To Reproduce Note that is the content of the in the above commit. note it also fails when overriding with however which is the current java15 version works. Expected behavior Successful build of SBT Additional context Non Debug output Notify maintainers Metadata Maintainer information:"}
{"_id":"q-en-nixpkgs-ebd2831434d99943b6832832d1f6a9939c68eec823ed66e670d0462e54632d63","text":"Describe the bug Nix installs a version of Chromium with Widevine on NixOS, despite not being enabled and goes through unfree license checks. To Reproduce Steps to reproduce the behavior: to system configuration in Chromium Expected behavior Widevine is not installed if not explicitly enabled. or Chromium trips nonfree license checks and forces user to allow unfree packages. Screenshots ! ! Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(patrycja): channels(root): nixpkgs: Maintainer information:\nChromium downloads the binaries itself and stores them in It looks like it's a known issue:\nOh, that's bad... :o Seems like Debian fixed this with a custom patch () but from your linked issue that doesn't seem really good either: It won't remove WidevineCdm automatically for existing users It won't be updated -vulnerabilities will never be fixed (which is very bad!) It might break other stuff: I don't think this applies to us though ( should manage it via the Nix store) Brief update: is also present on my system: ! Apparently not but looking at this version is already more than one year old! I cannot get any DRM demos to work so maybe the code will only download the binaries but not execute them (which is still bad but at least a bit better). https://shaka-player-;textlang=en-US;uilang=en-US;panel=HOME;build=compiled This even happens with chrome://settings/content/protectedContent set to \"Blocked\". It's disabled then but it will still be (re-)downloaded automatically.\nI've submitted this to upstream: Let's see what happens.\nIf you set enablewidevine you need a license to distribute it. See Over to the encrypted media folks to comment, but this sounds like WAI. Does Nix have a Widevine license? If Nix does not, and Google is unwilling to change how Chromium and Widevine are licensed together, the best course of action IMO would be to disable it in Nix's binaries (ie. if enableWidevine is false, enablewidevine is unset. If enableWidevine is true, enablewidevine is set and license is changed to unfree.) I understand that building Chromium is not something the average person can do on their toaster, but I don't think the group of people that need Widevine and are unwilling to use the official proprietary Chrome binaries is very large.\nI marked this as stale due to inactivity. &rarr;\nI am unable to reproduce this more than 3 years later on M124. Closing this, assuming this has been fixed a long long time ago.\nI can confirm this is still an issue. Followed instructions in OP.\nOkay, thanks, I managed to reproduce it now. Might have been a fluke or specific to my use of back when looked into this before. I want to clarify that this is not violation of the BSD-3 license uses. I do however understand that some prefer having not download additional unfree blobs.\nThere is now. It disables the automatic download and prevents previously downloaded blobs from being loaded and executed. With that PR, Widevine only works when going through the wrapper (). I hope this resolves the concerns raised in this issue here and makes both non-DRM and DRM users happy.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-77eb0b8a5f3028ea8d7ed25b29efba133d91dcc102a442455b7adeaed78aca7c","text":"Describe the bug After electron 12 came out, adding in home-manager succeeds, but trying to start the app fails mentioning a missing javascript function from an older version of electron. To Reproduce Steps to reproduce the behavior: to home-manager's app from terminal with Expected behavior Expected the application start as it had before. Additional context I was able to get it working locally by providing an overlay to home-manager: I would have created a PR myself, but I'm still new to nix and not sure entirely how to require a different electron version in the package derivation. Notify maintainers Metadata Please run and paste the result. system: host os: multi-user?: sandbox: version: channels(jon): nixpkgs: Maintainer information:\nthanks for catching this! Just fixed in the PR above."}
{"_id":"q-en-nixpkgs-03cb2c2cbbb6e88806ef0e63bfa4a0599c8dbaec1a9f9b035835ef6beba30541","text":"Describe the bug In , the path /usr/share/applications is hardcoded, the only way to change it would be to patch the package. However, from I see that the Debian package is used, so fixing this in the source code would require not to use a binary version (I guess). To Reproduce Steps to reproduce the behavior (tested with the package from the channel): the GUI client using the command the following error be thrown in the console: Expected behavior The autostart file is properly symlinked to the nix store package. Additional context I found a dirty fix for this issue, enable the preference and run: However, note that disabling this preference will replace the symlink with the one throwing the error. Notify maintainers Metadata Maintainer information:\nI always use sway and just started it that way on launch so I didn't notice. This will probably, as you said, need a change away from the pre-built package. Right now I'm quite busy, but I'll do it soon. Will tag you in the PR so you can test (as I can't...). Did the dir exist before changing the setting?\nHey, I'm working on the client daily, and I'm doing so on NixOS, so maybe I can help with this. Building the package might be a bit involved, since our is a bit involved ,and it relies on a rust toolchain installed by , although this could be worked around with some bash magic to just drop the argument. I personally just have a somewhat involved for building the daemon and then separate FSH environments for running electron builder, but that's because I'm also trying to build for Android and do weird debuggy stuff.\nYes it did since other packages created their own entries beforehand.\nThanks for the insight, So, is this an issue that might make sense to fix in the mullvad-vpn project itself? If not, could you please send me your so I have some guidance when building it?\nI must apologize for the hiatus, but life happened. I've been playing around with a somewhat cleaned up nix shell for building all the things for a proper nix package, but there's still some manual labor required. The shell that I use is somewhat similar to , but mine has far more cruft to also do android stuff. And this one could very well be slimmed down further too. The FSH environment for node is a vain attempt at making and work better, but to make things work, there's still some manual labor involved. The weird wrapping of is there because all the app's build scripts expect rust to be installed via rustup, whereas I use an overlay - so the nix script could be simplified. As long as the rust compiler is new enough, I think the app should build. To build the app, there are 6 parts that need to be built: for when the kernel module is not loaded, built via . Just executing that script on Linux should work just fine. the daemon, the CLI, problem report tool and OpenVPN plugin, should be buildable via , but the binaries end up being quite chonky, so they should all be stripped. Shell completions, the following snippet is copied verbatim from script. Download the relay list via Download the API address list via the electron GUI, built with an gulp that eventually invokes listed is basically a list of things that you must build or adjust so that they work in NixOS. Ideally, one would be able to just invoke , but due to  not building on nixOS (fails due to a node-gyp builder failing to find python) and producing an and a package, some patching would be required. It'd be great if a nix-npm wizard would apply some arcane knowledge to make build, otherwise it can just be removed from , as it's strictly a macOS dependency. Another change you'd probably want to make is to make make an unpacked build rather than producing RPMs and DEBs. The electron binary that's downloaded via will also not run. A dirty hack that I use to make it work is to just replace it with NixOS's one - , but this is not what the app developers are testing against, so there most certainly will be dragons. We build statically linked OpenVPN and Shadowsocks binaries, and those are found in  submodule. I assume these should be patchelf'ed so that they can be used on NixOS, but it seems like just using is not good enough, but doing what the current derivation does will most likely do the trick. These binaries may be substituted with NixOS native counterparts, but they wouldn't necessarily be equivalent - Mullvad's OpenVPN binaries have some patches that upstream are not interested in and we also try to build the minimal amount of features along with a minimal build of OpenSSL supporting only the ciphersuites that Mullvad servers use. I don't believe we have published the keys we use for signing commits currently, I'll see if we can make that happen. Currently, all the release commits are always signed by the same person - but this may change arbitrarily. After having written all of the above, I'm not entirely sure if making a native package would be worth the effort, given that I can't guarantee that any of the above won't change often. Maybe the best bet would be to make a minimal patch to get electron builder so that can just run. It seems to me that it's only electron/node related scripts that fail on nixOS: the GRPC buildscript fails because it expects to be able to call a downloaded binary from building fails when running probably some other electron quirk The commits are signed by individual developers, the keys are available , one should also make sure to check if the signatures are valid."}
{"_id":"q-en-nixpkgs-b11fb05ecc53210d4141776aa52b972953aaf8c383663a7181b5300c60392f7a","text":"Project description On some installations this simple text editor even has syntax highlightning, but it might require non-trivial amounts of work to get that part to work. Metadata homepage URL: source URL: Gitlab, , tag = or alternatively: latest current version: 1.2.2 license: MIT (according to the ebuild in the Gentoo repository) platforms: at least linux, but might also support other platforms with X11 support\nI marked this as stale due to inactivity. &rarr;\nI still need this."}
{"_id":"q-en-nixpkgs-683f2f198c5aea79e1bcc2b4383f1c4b428a18338368d83820b20c6636202f94","text":"The package hard-codes various X11 libraries as dependencies while making many other features optional, including GUIs. It would be nice to have an option to disable the X11 feature and the related . The use case I have in mind is installing a custom Vim on a remote server. Pulling X libraries, even without enabling a GUI, adds a few dozen megabytes for something essentially useless. An approach is doing something like this: but it doesn't feel very clean... Any thoughts from or ?\nI agree. It would have been nice if these X dependencies would have been to only if is or .\n/me whispers neovim~~~\nI'm all for testing new things and neovim is on my list. But according to , the neovim package pulls 428MiB of dependencies, half of which for Ruby, the other half for both Python 3 and Python 2 (seriously?). Is there a way, by any chance, to configure neovim to avoid this bloat ?\nadmittedly I've never tried to configure a mininmal version but one can unselect these ecosystems:\nSee ."}
{"_id":"q-en-nixpkgs-47bb524e6335bc7fc834a88f6621a352b28434c12f54659bd33132b972359433","text":"Project description a mining pool, maintains a fork of XMRig which adds the benefit of automatic algorithm switching. I have created following overlay which seems to work fine. The question is if anyone is interested in having it implemented as a separate package, replace the , or if it should be integrated as an option to a new XMRig systemd service (which would also be a nice-to-have, such as ) Notify Maintainers: Metadata homepage URL: source URL: license: gpl3 platforms: linux, darwin"}
{"_id":"q-en-nixpkgs-e85ed83079e394807d876239b11130c061efeb994ad81c53d5f7a0436d574335","text":"Describe the bug creates a defect as the block is defined two times. To Reproduce Steps to reproduce the behavior: Enable and Expected behavior and should be merged to one block. Additional context I'm on zfs. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(plumps): channels(root): nixpkgs: Maintainer information:\nI have a PR up to fix this"}
{"_id":"q-en-nixpkgs-e70f43ef7398225088613a3b14199ef5663d6d4da811d62d922eb8c4a037ab5a","text":"Describe the bug failed to build on master. To Reproduce Steps to reproduce the behavior: nixpkgs master Expected behavior It builds. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata Please run and paste the result. system: host os: multi-user?: sandbox: version: channels(root): channels(oxa): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-51c6a5be982b2309248fae6bcb8a312f003304962a41de41a971da8be78823c1","text":"Describe the bug Browser page fails to open if no installed on the system. To Reproduce Steps to reproduce the behavior: teleport using SAML (might need enterprise license) using tsh + SAML errors thrown, but the browser page doesn't open Expected behavior Browser page should open and ask for credentials. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata Please run and paste the result. Maintainer information:"}
{"_id":"q-en-nixpkgs-499a166c09c87a5c91ce8fe45424eb4dafbbd72ddb02154046275eb3fb584b47","text":"Describe the bug A clear and concise description of what the bug is. The and python scripts in need their interpreter to be patched (and the module available). To Reproduce Steps to reproduce the behavior: Expected behavior A clear and concise description of what you expected to happen. They should work :stuckouttongue:. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata Please run and paste the result. system: host os: multi-user?: sandbox: version: channels(bt1cn): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-6efbac3fd0923e08da6ac9ee93db879b76a819034a44be5869b98c9992f5485d","text":"Describe the bug wendy, one of the big-parallel build hosts used by , cannot successfully build This blocks advancing nixos-unstable. To Reproduce Steps to reproduce the behavior: wendy to build Observe :( or, more realistically: the attached CMOV16rm-, and run (a reduced commandline that produces the same assertion failure is ) Expected behavior A clear and concise description of what you expected to happen. Additional context The assertion that's tripped by this YAML file is: Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata Please run and paste the result. Maintainer information:\nis the CMOV16rm- (can't upload it here because it's not a supported GitHub filetype, apparently). To be more precise, the test that is failing is this one:\nThis has primarily become apparent because the configuration was changed in NixOS/nixos-org-configurations to make wendy only build big-parallel jobs. Since llvm is big-parallel, this means that wendy will now usually get selected for building llvm, since the load is otherwise pretty low. Previous successful builds of have been using one of the packet builders instead, which are using a newer CPU for which LLVM has a scheduling machine model.\nFWIW, this probably affects other LLVM versions as well, but they have current versions in the cache. If wendy rebuilds the other packages (e.g. stdenv bump or something), they'll likely start failing as well.\nIf you spin up a VM using the CPU profile that at least virt-manager calls \"OpteronG1\", then you can reproduce this using the LLVM tests. That's not quite the same as the CPU reported on wendy (OpteronG1: k8-sse3; wendy: amdfam10), but it exhibits the same problem.\nI don't know why, but wendy as disabled now.\nit's probably late for you (I hope I didn't confuse names), but you this issue some time ago.\nuh, same full name but this GitHub name is in .\nFiled upstream for this."}
{"_id":"q-en-nixpkgs-1cb22cf6968e91d3851337da11091b4805f2c17c92a5d5089b36dd74e8df423f","text":"Describe the bug when using , only gets set on linux. This seems to be done very much on purpose, so I don't know if there's some good reason for it: This causes problems for instance when cross-compiling to windows and then using pkgsStatic, some packages wrongly think they are not compiling statically. I can't remember what packages I've encountered this in, and right now I'm encountering it in a change that I'm trying to make, not something that already exists. To Reproduce Steps to reproduce the behavior: Expected behavior would be true on all platforms if using Notify maintainers seems to have this code / created Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(qolen): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-acda0064354ca6fd0d9b699d200b641b57f4417ffbcf3ea3b56abb7ab45e6155","text":"Describe the bug On current nixos-unstable (), fails to build To Reproduce Steps to reproduce the behavior: -shell -p ophis Notify maintainers"}
{"_id":"q-en-nixpkgs-ee0db80ec4d0ac01707f19e46b738fcfb8aadba0786484cd52b01d93896101da","text":"Describe the bug The Airsonic service fails to start. To Reproduce Steps to reproduce the behavior: = true; to your config. reveals the following error: Cannot load configuration class: which causes the application to enter a restart loop. <details<summaryError trace</summary<p</p</detailsExpected behavior It doesn't error, and instead runs successfully. Additional context It doesn't seem to be due to an Airsonic update that this started to fail, so my guess is that something got changed with the JRE. I think Airsonic will only work with JDK 8 (see: ), but I'm not very familiar with how Java works, so I'm not sure. Notify maintainers Metadata Fails on both of the following systems: Maintainer information:"}
{"_id":"q-en-nixpkgs-69f6b2d35ce7f68ade89129b7d0c6df32307af9063c10fc159e285d71eec3544","text":"Describe the bug Unbound allows the \"include\" directive to appear more than once. : However the type definition of is too strict. This code raises an error: Notify maintainers Metadata Maintainer information:\nCould you provide me with the value you set the option to?\nAn example has been provided above. Simply set to a list of string. The service will fail, but it should work. You can try\nAh yes indeed, sorry I didn't see that. I'll open a PR that fixes this."}
{"_id":"q-en-nixpkgs-1056bea3e6a56690fbd716fd22ddafe958afdf8fa0ffd5e3aebcfb340d93093b","text":"Project description Add Youtube Music Desktop player Metadata homepage URL: source URL: license: Creative Commons Zero v1.0 Universal platforms: linux, MacOS Similar application to Google Play Music Desktop:"}
{"_id":"q-en-nixpkgs-29800f76fc20e14a82f535158e459b62937b14408bf9f659e0dc587be5675d67","text":"On NixOS 20.09, steps to reproduce the behavior: setting up wireguard, change to another port still shows the old port:\nThis seems to be fixed by my PR"}
{"_id":"q-en-nixpkgs-f7eebc3a2cf59fa1b877c1e72aa3cddaa58e943112fdb8e811c501e06c0652dc","text":"Describe the bug VAAPI records repeated frames (see upstream bug report). upstream fix: To Reproduce Steps to reproduce the behavior: ... open OBS ... try recording using VAAPI (possible AMD hardware required?) ... watch replay Expected behavior full FPS. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Maintainer information:\nAs for 21.0, the fix contained in upstream release = 21.0.2 so e.g. PR\nSadly a look at that thread, it's blocked because something broke, and they're not pushing it to 21.05, so this bug is going to exist until someone figures this out the regression, this patch is only one line of code, maybe it can be supplied as a patch until they get 21.0.3 working? Is there a way to build a patched mesa without rebuilding my entire desktop stack? It's currently stupidly slow on my computer when doing a packageOverride.\nThis will not be the only bug people will have to deal with on 21.05 unless we backport a more recent mesa (=21.1). Alternative we could also start to includes as well. Without it, this will lead to crashes in wine (incl. steam proton) when on some hardware ( for reference). While one can set as a workaround, it is hardly ideal and difficult to diagnose from within steam alone. Would it be possible to backport the mesa update currently in staging-next () once it is merged?\nI don't know... this is hard. Maybe after a version has been a couple weeks in , we might be sufficiently confident that it won't make things worse overall.\nI'm okay with backporting it, but probably as its own job set, just to get a better idea of regressions.\nI'm not afraid of build-time regressions in this case. Those would be easy (to find). And mesa isn't such a big rebuild now anyway.\nIf it's been on unstable for a while, I guess I'm okay with it as well\nI guess we can close this now, I haven't noticed any problems recently."}
{"_id":"q-en-nixpkgs-c1d01c6a15b8d34ec533b06fdc04843931ebefb8b1f0acad76ec12ec6398fda7","text":"Drone 2.0 has been released: <!-- Note that these are hard requirements --- name: <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 1.10.0 desired version: 2.0.0 Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-24108837e07d69d2a3037efb64ada296d8992e63f63736d4ffab56cc84a37388","text":"[ ] It's likely that kodi/xbmc remains vulnerable. [ ] Audacity on 15.09 is older, so the patch used on master doesn't apply and wouldn't be enough anyway.\nWell, the two minor issues above remain, so I'll leave this open ATM.\nClosing as 15.09 has now been superseded by 16.03 and kodi is using ffmpeg 2.8.6 since"}
{"_id":"q-en-nixpkgs-8a12b75ef72fb2011bab60e7e0afdd438e7f373d3034763949e4cff52391750e","text":"Describe the bug On NixOS and some other Linux distributions like Fedora is world-read-writable: This is not the case on other systems like Ubuntu and others from the Debian family, which instruct their users to get access to by adding user accounts to the group: When building a derivation that uses , like for example NixOS Tests or on such systems with Multi-User Nix and KVM installed, the build fails with an error message similar to this when booting the VM: To fix this users might try to add the users to the group, which has no effect because of the build sandbox. The only effective fix seems to be changing the permissions for to be world-read-writable: This fix is difficult to find for users that are not already very familiar with KVM (like me). Proposed fix Nix should detect the situation where a build is run in the sandbox and the permissions for are going to lead to a failure and emit a warning or build error that points people towards a working fix they can manually run. Potentially it would even be possible/desirable for the build to proceed without KVM/hardware acceleration, but I'm not sure because of the terrible performance impact. To Reproduce Steps to reproduce the behavior: a system that supports hardware virtualization ubuntu or another affected distro KVM (see for example ) Multi-User Nix a derivation that uses You can also manually change the permissions to to reproduce the issue. Expected behavior A more user-friendly failure mode, like what I wrote under proposed fix. Additional context There are also documented examples of other users running into this: and maybe also There is some relevant discussion here about the security impact of that permission change: Notify maintainers There are a lot of other people in git blame. I you two for now. Hope that's OK. Maintainer information:\nVMs already proceed without hardware acceleration if KVM is not installed at all. I tested this with NixOS Tests on with NixOS. At least for that case the relevant code I found is this: I will submit a PR which fixes this check so that VMs also proceed without KVM in the scenario described above. I am not 100 % sure if that is the best fix and if it should close this issue, because firstly I think it would be good to make users aware when KVM is not enabled, when the hardware/software support would allow it: because it's not installed or because the permissions are broken. What would be a good way to emit such warnings? Secondly I am also a bit confused about there being flags for hardware acceleration here as well: This makes it seem like is treated inconsistently with other architectures and I don't understand why."}
{"_id":"q-en-nixpkgs-ccf5bf3669d5b0707518cde5cbf36e63aff068be446521ec457e060408f9fc22","text":", CVSSv3=7.8 (nixos-20.09, nixos-unstable) Scanned versions: nixos-20.09: ; nixos-unstable: . Cc Cc\n09 is still vulnerable\n09 is EOL"}
{"_id":"q-en-nixpkgs-437ee1aa65426e2d843484fa3f9f69c776314ce016ff6824c42ee22d498f23cb","text":"Describe the bug Papermc crashes due to an incompatible Java version To Reproduce Steps to reproduce the behavior: Set Log Excerpt Additional context The vanilla package runs without issues. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata system: host os: multi-user?: sandbox: version: channels(johannes): channels(root): nixpkgs: Maintainer information:\nI couldn't reproduce with this test driver : <details</details\nI also get this. I think papermc starts supporting java 16 with build 668 (at least it seems to start working for me after that build), but the current build in nixpkgs is 488 building the latest build (771) works for me. the test you defined doesn't wait long enough to actually start the server Which I think you need to run with since it requires networking to download the vanilla minecraft jar it will then patch. Now the test fails as expected complaining about the java version. But if you overwrite the papermc package with: the test will succeed. (btw is there a smarter way to overwrite these variables?) Metadata: system: host os: multi-user?: sandbox: version: channels(root): channels(kaine): nixpkgs:\nSame thing happens with multimc and the new version"}
{"_id":"q-en-nixpkgs-969f85afd78ca0fa0455a0cc5c6b18825b95597ab6d2d7ff8872c4f323c9e7c8","text":"Project description A system for automatically configuring mutt and isync with a simple interface and safe passwords Metadata homepage URL: source URL: license: gpl3 platforms: unix, linux, darwin"}
{"_id":"q-en-nixpkgs-490c82dc7fcbf2e56a85eef5bdc252fd2c46d825989f2b82518a4b2a414a7dce","text":"Project description The terminal is already packaged in nixpkgs. As foot uses its own terminfo, this terminfo is needed on any remote servers you connect to using foot. It would be useful to have this terminfo packaged as its own output as with other terminals: alacritty, kitty, rxvt-unicode to name a few. ping foot maintainer for feedback Metadata homepage URL: source URL: license: mit platforms: linux\nThat is a very good idea! Unfortunately it doesn't seem to be possible to specify a separate install directory for foot's terminfo so far (it is installed to along with the license etc.). I'll open an upstream ticket, they are usually very helpful, and then we'll be hopefully be able to set this pretty soon."}
{"_id":"q-en-nixpkgs-7f205a66beb2dee5bfecda0c344c36947337f5119d37a0f0bb9b076194c22930","text":"Describe the bug Building or fails with: To Reproduce Steps to reproduce the behavior: until approximately 70% (sometimes a few percent earlier) Additional context Compilation log for with only : Compilation log for with only : I had this problem with Sierra with XCode 9.2 also. I just upgraded to Mojave 10.14.6 and XCode 11.13.1. Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata system: host os: multi-user?: sandbox: version: channels(peterhil): channels(root): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-a3364ba1f1f39b31b8a84aafbbd82fd29a0b0fa5f0352559c56081cb4d3a331d","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on github to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: nixos.poppler <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 21.05.0 desired version: 21.06.01 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR. The current version of is causing evince to crash frequently:"}
{"_id":"q-en-nixpkgs-a652003c6c3483e8a85b9232e9e6e58ea1a22ac7b9accdec1748d99cbe3f4f37","text":"Currently, we first build a derivation containing the unpacked Chromium source code, which is then passed to the actual Chrome build. This is very inefficient: the unpacked source tree is 1.8 GB and ~85K files. Copying these between build machines can take a very long time: for instance, the is currently showing two builds (in the copying state) which have already taken 22 hours. So it would be better to do whatever does in the derivation.\ndidn't we do this because IO was too slow for unpacking in NixOS tests?\nI don't think the tests need the source code.\nIs there another way to avoid copying source derivations between machines? I think the source derivation is good to keep, I want to have QtWebEngine which uses chromium, would be nice to use chromium source and all patches that is used. Would this still be possible (in a easy way) if chromium-source is removed?\nThe derivation could create a (compressed) tarball. That would help a lot since then Hydra won't have to do a huge amount of random access I/O reading/writing lots of small files.\nThe reason I've done it that way was to separate bundled dependencies and the sandbox from the main source, so for example if we want to build just the sandbox, we don't need to extract the full Chromium source. But I'll look into it so we can find a better solution, right now it's not working as I had in mind back then, because the source is copied into the temporary build directory, but a better solution would be to not copy anything at all.\nI don't think you can re-use for this, as Qt uses a rather heavily modified version of the source.\nGreat point. I will add too that even if it was vanilla, chrome source moves far too quickly to be a dependency of a qt component. QtWebEngine is; I suspect using their own fork of the blink fork of webcore (webkit.) (webengine) If we look at the chromium release schedule they are trying to hit a release almost every month. If someone is quick (Before 19th of this month) they could apply for a gsoc to possibly get this chromium work done (worth a try?)\nWith S3 and decentralized hydra, we should really combine the source back. Currently this is blocking a channel update:\nWorking on this right now.\nI've also the relevant changes to .\nThere's reference to the old path of the file on line 33 of this file:"}
{"_id":"q-en-nixpkgs-dec12f4d351ece766ccbd29f574f8779c22206ff6633bbbc90a24fc9143d3b70","text":"Project description The driver supports Realtek 8852AE, which is a wifi card for some new laptops. Without the driver, the wireless interface would be missing for those computers. (My new Lenovo laptop uses this card and I cannot connect to WiFI currently... I'm not entirely sure how to package this) Metadata homepage URL: source URL: license: GPA 2.0 platforms: Linux"}
{"_id":"q-en-nixpkgs-35122fbe11d31e3dd233d7e2e88ec505fa958ea47ae86d41464f94bebe7f4624","text":"Describe the bug With the latest kernel and the latest non-beta nvidia driver, the drivers don't compile: To Reproduce Steps to reproduce the behavior: unstable the latest linux kernel nvidia as a drivers Expected behavior It should compile. Notify maintainers Latest person who touched it: Metadata\nI can reproduce this issue, albeit I am using stable channel with the latest kernel.\ntl;dr for everyone facing this: should serve as a temporary workaround\nI think we should implement some kind of kernel version lock for the nvidia driver to prevent such occurrences in the future."}
{"_id":"q-en-nixpkgs-2556965b3882f168be5dafa79bf925433f9a54fcc84387044efdec3a416293b7","text":"I get the following output from :\nThanks, So fast. :)\nThis one was rather easy :-)"}
{"_id":"q-en-nixpkgs-c97b6fc8ac269316e4437f96a61c176900e8ed654077d9ed9ded1f4dd4e55538","text":"Describe the bug The cinnamon-menu-editor doesn't open the cinnamon-desktop-editor, if I want to change the 'Properties' of a menu Entry. To Reproduce https://user- Expected behavior https://user- Additional context Here is the cinnamon-desktop-editor source: Notify maintainers thanks for making cinnamon available in NixOS! :) Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(linux): nixpkgs: Maintainer information:\nOpen the menu editor through the terminal and edit the properties then send the output\nI open cinnamon-menu-editor in Terminal, select an entry, click \"Properties\": This is in the launcher- :\nI marked this as stale due to inactivity. &rarr;\nThis patch fixes the issue, if nobody wish to investigate on this issue I will go ahead and commit this change on next cinnamon update. Edit: Ah I probably don't need to set dontWrapPythonPrograms Edit 2: Non English user can still encounter some l10n issues like translated strings are not modified, we will track the issue in"}
{"_id":"q-en-nixpkgs-f372e129959372975f314d0d2a4149c963fcbea2a6b81a8bea8154c2aa05a241","text":"Describe the bug Neo4j does not start because it cannot locate a Java Runtime. To Reproduce On my system, I installed with: Then, running shows: \"The operation couldn’t be completed. Unable to locate a Java Runtime.\" Expected behavior Neo4j should start. Additional context Note that shows Note the line that sets to . Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata Please run and paste the result. Maintainer information:"}
{"_id":"q-en-nixpkgs-1b96ef8969f8b3064314baf27f7c5ec057343f60f02778062239bacf146388da","text":"Describe the bug A clear and concise description of what the bug is. To Reproduce Steps to reproduce the behavior: to chromium 91.0.4472.164 (current on unstable channel) chromium that some area where there should be text do not render, this also happens in the developer console or at chrome://gpu/ for example. Expected behavior No error in font rendering Screenshots ! or ! Additional informations Works on some pages, like or teams, when directly accessing a meeting. Same behavior observed by a colleague running qutebrowser: ! (this is ) ^. ; ; ;,- ,- ,-\"\"\"   ^^ ,-; ; ,- / ;; ,^ / ,^ ,-\" - qutebrowser v2.3.0 Git commit: Backend: QtWebEngine 5.15.3, based on Chromium 87.0.4280.144 Qt: 5.15.2 CPython: 3.9.5 PyQt: 5.15.4 sip: 6.1.0.dev2104271705 colorama: no jinja2: 3.0.1 pygments: 2.9.0 yaml: 5.4.1 adblock: 0.5.0 PyQt5.QtWebEngineWidgets: yes PyQt5.QtWebEngine: 5.15.4 PyQt5.QtWebKitWidgets: no 2.8.335 (/nix/store/z5n6n43w80lh2p2pfp694q38chlrmnl0-pdfjs-) sqlite: 3.35.5 QtNetwork SSL: OpenSSL 1.1.1k 25 Mar 2021 Style: QFusionStyle Platform plugin: xcb OpenGL: Intel, 4.6 (Compatibility Profile) Mesa 21.1.5 Platform: Linux-5.10.48-x86_64-with-glibc2.33, 64bit Linux distribution: NixOS 21.11 (Porcupine) (nixos) Frozen: False Imported from /nix/store/42sm4xmn4z74m19q7n70b7bgikd9qcy9-qutebrowser-2.3.0/lib/python3.9/site-packages/qutebrowser Using Python from /nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/bin/python3.9 Qt library executable path: /nix/store/i1gvlgf0g8scyc9rqk9gcz0pvwnvy6sp-qtbase-5.15.2/libexec, data path: /nix/store/i1gvlgf0g8scyc9rqk9gcz0pvwnvy6sp-qtbase-5.15.2 yaml attribute: chromium module:\nRolling back to chromium v91.0.4472.114 fixes the problem:\nI can confirm this issue. All browsers based on seem to be affected. Qutebrowser definitely is.\nI can also confirm it (noticed it a few hours ago) but didn't have time to investigate so far. A good revision is and a bad revision so something between those two revisions must've caused this regression (it's unrelated to Chromium updates). I suspect either or maybe .\nI'm personally experiencing this in both qutebrowser and Chromium (but Chromium works with on the command line). From what I can tell, this does seem to be caused by the glibc 2.33 update: Chromium bug tracker issue QtWebEngine bug tracker issue Fix in Chromium (this one got reverted) (in version 94, the current dev channel version) Fedora patches for Chromium and QtWebEngine Arch patches for Chromium and QtWebEngine\n(qtwebengine maintainer): Could you look into the fix for QtWebEngine (see )? cc (qutebrowser maintainers) Regarding Chromium: should hopefully fix this regression:"}
{"_id":"q-en-nixpkgs-4825b3f485de405ca8b15f44be3c73a96f1f0e2ee9eef5cf72493a6b8123b774","text":"Describe the bug In the 21.05 VirtualBox OVA you can download from there is no disk space available, making it impossible to use for almost anything: To Reproduce Steps to reproduce the behavior: the VirtualBox OVA from , the instructions on above page to import and start it (although the instructions are a bit outdated too, no need to log in manually) in a Konsole window Expected behavior I expect to have disk space. Here's the same for a NixOS 20.09 OVA: Screenshots ! Notify maintainers Metadata Bad nixpkgs version (21.05): Good nixpkgs version (20.09): Maintainer information:\nhmm, well we can add a new option to the for the additional disk space in the image. We then could pass the value to from\nI think that's setting the disk image size. Which is given here: I haven't had time to take a look, I wonder what looks like; is it a situation where fails?\nRight, so assumption seems more right. Good to know."}
{"_id":"q-en-nixpkgs-f6604abd5d2b433f03b67561d682fbc391d8a7afde81da76b56aa7a66831f4ba","text":"Helvum build is broken on unstable Steps to reproduce the behavior: master helvum Build suceeds. Since it's pipewire-sys crate that fails, I suspect it was a pipewire upgrade that broke it Maintainer information:\nTrying to build latest helvum master, as it's now been updated to use pipewire from Will post updates.\nAlright, it fails even with the latest master, so I assume it's a pipewire error. is the maintainer of the PipeWire package and the leader of the PipeWire efforts on Nix/OS so they may know more about this. In the meantime, helvum will remain broken :(\nIs it the same error? Could be that the bindings need an upgrade to support the newer version of pipewire because of changed APIs. Then that would be a bug in the pipewire-sys crate.\nYes, it's exactly the same error. You're right, I'll look into pipewire-related issues on GitLab to try and see what's going on.\nI found , which means we have to wait a bit until that gets merged.\nThe issue links to this pull request: Not sure how it's with crates, but the dependency could be patched using this patch: (This is the alternative to waiting until that's resolved)\nSure, I'll get to work then.\nStrictly speaking I'm not officially maintaining the pipewire package, but I guess I've effectively become the de facto maintainer anyway.\nSorry for mispinging you then!\nNo problem. I originally just wanted to get some progress and organization in the messy packaging situation. I hope that the official maintainers and autoupdate bot can take over again once upstream stabilizes more. I'm not actually sure who exactly the official maintainers would be, though. The package just says \"freedesktop maintainers\". EDIT: and I did end up following upstream development pretty closely so pinging me was a very reasonable guess.\nas it currently stands, I don't think it's possible to override the dependency with a nix-built one. Though I have made a nix derivation for it, there's no way to propagate it to using . In turn, I'll wait and see what Ryuukyu does upstream.\nI've fixed pipewire-rs and helvum and made a new release, so it should now build again once you update the package.\n! Can confirm it works. Thanks"}
{"_id":"q-en-nixpkgs-568a38cbc0073c98b638220375a46ea7303bd0ed033d44731bb76562410fb565","text":"The calibre derivation fails to build on master () Steps to reproduce the behavior: flakes on my nixos I do calibre is not yet build on hydra it has to compile locally fails, see log below compile or download calibre an run it <details<summary<codenix log /nix/store/yr3aa75ncgmcbmacwl2b28l04k1gr29p-calibre-</code</summary<pre { \"action\": \"setPhase\", \"phase\": \"qmakePrePhase\" } qmakePrePhase { \"action\": \"setPhase\", \"phase\": \"qtPreHook\" } qtPreHook { \"action\": \"setPhase\", \"phase\": \"unpackPhase\" } unpacking sources unpacking source archive /nix/store/qkps06f9m4p4sv5ifyxql689y05mfm0q-calibre- source root is calibre-5.17.0 setting SOURCEDATEEPOCH to timestamp of file calibre-5.17.0/man-pages/vi/man1/web2disk.1 { \"action\": \"setPhase\", \"phase\": \"patchPhase\" } patching sources applying patch /nix/store/1442hsfaj6iw5s08h13w6s1v6zazn07s- patching file Hunk succeeded at 30 with fuzz 2 (offset 2 lines). applying patch /nix/store/l042hpvp785l8424slj5nqrncrh6jmzm- patching file Hunk succeeded at 59 (offset 22 lines). applying patch /nix/store/qdaan22nwhppgj0df96gs02wzg006x3s- patching file Hunk succeeded at 69 (offset 25 lines). { \"action\": \"setPhase\", \"phase\": \"configurePhase\" } configuring no configure script, doing nothing { \"action\": \"setPhase\", \"phase\": \"buildPhase\" } building no Makefile, doing nothing { \"action\": \"setPhase\", \"phase\": \"glibPreInstallPhase\" } glibPreInstallPhase { \"action\": \"setPhase\", \"phase\": \"glibPreInstallPhase\" } glibPreInstallPhase { \"action\": \"setPhase\", \"phase\": \"installPhase\" } installing Running build Building 24 extensions Compiling 35 files... g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/qdy5zcdxx836cn4iiwz19kz7i57pqnjr-hunspell-1.7.0-dev/include/hunspell -c /build/calibre- -o /build/calibre-5.17.0/build/objects/hunspell/hunspellwrapper.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c99 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/utils/hyphenation/hyphen.c -o /build/calibre-5.17.0/build/objects/hyphen/hyphen.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c99 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/unicodenames/unicodenames.c -o /build/calibre-5.17.0/build/objects/unicodenames/unicodenames.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/utils/speedup.c -o /build/calibre-5.17.0/build/objects/speedup/speedup.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/utils/certgen.c -o /build/calibre-5.17.0/build/objects/certgen/certgen.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/gui2/tweakbook/editor/syntax/html.c -o /build/calibre-5.17.0/build/objects/htmlsyntaxhighlighter/html.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/tinycss/tokenizer.c -o /build/calibre-5.17.0/build/objects/tokenizer/tokenizer.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/gui2/tweakbook/diff/patiencediffc.c -o /build/calibre-5.17.0/build/objects/patiencediffc/patiencediffc.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/utils/icu.c -o /build/calibre-5.17.0/build/objects/icu/icu.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/utils/matcher.c -o /build/calibre-5.17.0/build/objects/matcher/matcher.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/library/sqlitecustom.c -o /build/calibre-5.17.0/build/objects/sqlitecustom/sqlitecustom.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/lzx -c /build/calibre-5.17.0/src/calibre/utils/lzx/compressor.c -o /build/calibre-5.17.0/build/objects/lzx/compressor.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/lzx -c /build/calibre-5.17.0/src/calibre/utils/lzx/lzxmodule.c -o /build/calibre-5.17.0/build/objects/lzx/lzxmodule.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/lzx -c /build/calibre-5.17.0/src/calibre/utils/lzx/lzc.c -o /build/calibre-5.17.0/build/objects/lzx/lzc.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/lzx -c /build/calibre-5.17.0/src/calibre/utils/lzx/lzxd.c -o /build/calibre-5.17.0/build/objects/lzx/lzxd.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/lzx -c /build/calibre-5.17.0/src/calibre/utils/lzx/lzxc.c -o /build/calibre-5.17.0/build/objects/lzx/lzxc.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -c /build/calibre- -o /build/calibre-5.17.0/build/objects/freetype/freetype.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/msdes -c /build/calibre-5.17.0/src/calibre/utils/msdes/msdesmodule.c -o /build/calibre-5.17.0/build/objects/msdes/msdesmodule.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/calibre/utils/msdes -c /build/calibre-5.17.0/src/calibre/utils/msdes/des.c -o /build/calibre-5.17.0/build/objects/msdes/des.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/ebooks/compression/palmdoc.c -o /build/calibre-5.17.0/build/objects/cPalmdoc/palmdoc.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/ebooks/djvu/bzzdecoder.c -o /build/calibre-5.17.0/build/objects/bzzdec/bzzdecoder.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/output.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/impose.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/images.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/utils.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/doc.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/outlines.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/outline.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/fonts.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include/podofo -c /build/calibre- -o /build/calibre-5.17.0/build/objects/podofo/podofo.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++11 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -c /build/calibre- -o /build/calibre-5.17.0/build/objects/htmlasjson/htmlasjson.o g++ -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -std=c++14 -DCALIBREMODINITFUNC=extern \"C\" attribute ((visibility (\"default\"))) PyObject -I/build/calibre-5.17.0/src/perfect-hashing -c /build/calibre- -o /build/calibre-5.17.0/build/objects/fastcsstransform/fastcsstransform.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/devices/libusb/libusb.c -o /build/calibre-5.17.0/build/objects/libusb/libusb.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/devices/mtp/unix/libmtp.c -o /build/calibre-5.17.0/build/objects/libmtp/libmtp.o gcc -Wall -DNDEBUG -fno-strict-aliasing -pipe -fPIC -fvisibility=hidden -pthread -fdiagnostics-color=always -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -O3 -DCALIBREMODINITFUNC= attribute ((visibility (\"default\"))) PyObject -c /build/calibre-5.17.0/src/calibre/devices/mtp/unix/devices.c -o /build/calibre-5.17.0/build/objects/libmtp/devices.o Linking 21 files... g++ /build/calibre-5.17.0/build/objects/hunspell/hunspellwrapper.o -o /build/calibre- -Wall -shared -lpython3.9 -L/nix/store/52l4k58ycwq1xc8z5wsk379q2ay0n7bb-hunspell-1.7.0/lib -lhunspell-1.7 gcc /build/calibre-5.17.0/build/objects/hyphen/hyphen.o -o /build/calibre- -Wall -shared -lpython3.9 -lhyphen gcc /build/calibre-5.17.0/build/objects/unicodenames/unicodenames.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/speedup/speedup.o -o /build/calibre- -Wall -shared -lpython3.9 -lm gcc /build/calibre-5.17.0/build/objects/certgen/certgen.o -o /build/calibre- -Wall -shared -lpython3.9 -lcrypto gcc /build/calibre-5.17.0/build/objects/htmlsyntaxhighlighter/html.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/tokenizer/tokenizer.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/patiencediffc/patiencediffc.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/icu/icu.o -o /build/calibre- -Wall -shared -lpython3.9 -licudata -licui18n -licuuc -licuio gcc /build/calibre-5.17.0/build/objects/matcher/matcher.o -o /build/calibre- -Wall -shared -lpython3.9 -licudata -licui18n -licuuc -licuio gcc /build/calibre-5.17.0/build/objects/sqlitecustom/sqlitecustom.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/lzx/compressor.o /build/calibre-5.17.0/build/objects/lzx/lzxmodule.o /build/calibre-5.17.0/build/objects/lzx/lzc.o /build/calibre-5.17.0/build/objects/lzx/lzxd.o /build/calibre-5.17.0/build/objects/lzx/lzxc.o -o /build/calibre- -Wall -shared -lpython3.9 g++ /build/calibre-5.17.0/build/objects/freetype/freetype.o -o /build/calibre- -Wall -shared -lpython3.9 -L/nix/store/d7gldi24058s3jk0ssn4ghr7mj8w4nca-freetype-2.10.4/lib -lfreetype gcc /build/calibre-5.17.0/build/objects/msdes/msdesmodule.o /build/calibre-5.17.0/build/objects/msdes/des.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/cPalmdoc/palmdoc.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/bzzdec/bzzdecoder.o -o /build/calibre- -Wall -shared -lpython3.9 g++ /build/calibre-5.17.0/build/objects/podofo/output.o /build/calibre-5.17.0/build/objects/podofo/impose.o /build/calibre-5.17.0/build/objects/podofo/images.o /build/calibre-5.17.0/build/objects/podofo/utils.o /build/calibre-5.17.0/build/objects/podofo/doc.o /build/calibre-5.17.0/build/objects/podofo/outlines.o /build/calibre-5.17.0/build/objects/podofo/outline.o /build/calibre-5.17.0/build/objects/podofo/fonts.o /build/calibre-5.17.0/build/objects/podofo/podofo.o -o /build/calibre- -Wall -shared -lpython3.9 -L/nix/store/np4f7wvx6kpp4i3yxz1qmzmzwdnq2d0l-podofo-0.9.7-lib/lib -lpodofo g++ /build/calibre-5.17.0/build/objects/htmlasjson/htmlasjson.o -o /build/calibre- -Wall -shared -lpython3.9 g++ /build/calibre-5.17.0/build/objects/fastcsstransform/fastcsstransform.o -o /build/calibre- -Wall -shared -lpython3.9 gcc /build/calibre-5.17.0/build/objects/libusb/libusb.o -o /build/calibre- -Wall -shared -lpython3.9 -lusb-1.0 gcc /build/calibre-5.17.0/build/objects/libmtp/libmtp.o /build/calibre-5.17.0/build/objects/libmtp/devices.o -o /build/calibre- -Wall -shared -lpython3.9 -lmtp SIPing 3 files... /nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/bin/python3.9 -c import os; ('/build/calibre-5.17.0/build/pyqt/pictureflow'); from import main; main(); --verbose --no-make --qmake /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake Querying qmake about your Qt installation... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -query These bindings will be built: pictureflow. Generating the pictureflow bindings... Generating the .pro file for the pictureflow module... Generating the top-level .pro file... Generating the Makefiles... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -recursive Info: creating stash file /build/calibre- Reading /build/calibre- The project has been built. /nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/bin/python3.9 -c import os; ('/build/calibre-5.17.0/build/pyqt/progressindicator'); from import main; main(); --verbose --no-make --qmake /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake Querying qmake about your Qt installation... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -query These bindings will be built: pictureflow. Generating the pictureflow bindings... Generating the .pro file for the progressindicator module... Generating the top-level .pro file... Generating the Makefiles... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -recursive Info: creating stash file /build/calibre- Reading /build/calibre- The project has been built. /nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/bin/python3.9 -c import os; ('/build/calibre-5.17.0/build/pyqt/imageops'); from import main; main(); --verbose --no-make --qmake /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake Querying qmake about your Qt installation... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -query These bindings will be built: pictureflow. Generating the pictureflow bindings... Generating the .pro file for the imageops module... Generating the top-level .pro file... Generating the Makefiles... /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -recursive Info: creating stash file /build/calibre- Reading /build/calibre- The project has been built. cd pictureflow/ && ( test -e Makefile /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -o Makefile /build/calibre- ) && make -f Makefile make[1]: Entering directory '/build/calibre-5.17.0/build/pyqt/pictureflow/build/pictureflow' g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sippictureflowcmodule.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sippictureflowPictureFlow.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sippictureflowFlowImages.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o pictureflow.o g++ -pipe -O2 -fno-exceptions -Wall -Wextra -dM -E -o mocpredefs.h /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2- /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/moc -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB --include /build/calibre-5.17.0/build/pyqt/pictureflow/build/pictureflow/mocpredefs.h -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -I/build/calibre-5.17.0/build/pyqt/pictureflow/build/pictureflow -I/build/calibre-5.17.0/build/pyqt/pictureflow/build/pictureflow -I/build/calibre-5.17.0/build/pyqt/pictureflow/build -I/build/calibre-5.17.0/src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include -I/nix/store/1yfjzqywgdrllq3cxvr5b7bkaa12ana4-libxml2-2.9.12-dev/include -I/nix/store/gxhfbfaqkh5pna8ap5xajjzlcj2dcp4y-zlib-1.2.11-dev/include -I/nix/store/wfwyqavhqwnqrjgq1jyhg2g9g6ya9awf-libxslt-1.1.34-dev/include -I/nix/store/qsl4jjnr5ji16nl636kdj24iiasj1ws4-openssl-1.1.1k-dev/include -I/nix/store/fiacxkszwy3gah3jx8c2g6rby5sj5g3c-sqlite-3.35.5-dev/include -I/nix/store/8z6xcnagxrgzs6hm5n7rvw7fqffgdlva-harfbuzz-2.8.1-dev/include -I/nix/store/6iwy5bxv8p64fizdd9m6277x8sbgaqq4-graphite2-1.3.14/include -I/nix/store/gn3qqpj4zi49pyjg5mf30aks07k49650-icu4c-69.1-dev/include -I/nix/store/qk9wf7rxzpi3b2yjsrglfy2ji482yscx-libjpeg-turbo-2.1.0-dev/include -I/nix/store/4kiqgcp4zjg25hpwm8mpv12hfi1qwiba-libpng-apng-1.6.37-dev/include -I/nix/store/63s3m6ifs7lcfx8icz7gz59qzabkww6z-pcre2-10.36-dev/include -I/nix/store/4ipg90ns4kx9p5qpwj0f6d6ysgf9x9lx-dbus-1.12.20-dev/include -I/nix/store/3lr9nqspakav45w936rb4y5q4hlps87b-expat-2.4.1-dev/include -I/nix/store/6l8vy0rpikcw0j1xw67qf983sqpxcfzp-glib-2.68.3-dev/include -I/nix/store/dxlyb8r1nxvjmak635rm1r654ig69z02-libffi-3.3-dev/include -I/nix/store/vyy5f2lnkq1b3r7rr79x1hwwcj2di3w6-gettext-0.21/include -I/nix/store/sxxkqm7smpdxwi985a047mvxdmgaw2bf-glibc-iconv-2.33/include -I/nix/store/hd68r4arz1smy4skafkvwzw5q2r61hk8-systemd-247.6-dev/include -I/nix/store/0fqvkh0lz98bn2zwiarfij48dzblg9r8-fontconfig-2.13.94-dev/include -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include -I/nix/store/lsfybn0mjjdx8qa6zf9n98xq2vj5fvy7-bzip2-1.0.6.0.2-dev/include -I/nix/store/s8dw25jfsqrx40q5m8rx8g4r48nnvbkb-libX11-1.7.2-dev/include -I/nix/store/z8k375ymlx56l6mhm6y37sfb8s06mxd0-xorgproto-2021.4/include -I/nix/store/mjcc6qjh7r729m3cnl8yh61pg07hqpxr-libxcb-1.14-dev/include -I/nix/store/brda65m3j4kym70mmvhn1j90gaysvpcs-libXcomposite-0.4.5-dev/include -I/nix/store/42ds8bnqg6sghqq6q4hm0f8xgryhk7q2-libXfixes-6.0.0-dev/include -I/nix/store/ibmnb203lx07sx840kqyjzjyms6kjplh-libXext-1.3.4-dev/include -I/nix/store/iyymggidpjja58ljzwl2vfbfyr8a8j4x-libXau-1.0.9-dev/include -I/nix/store/0b5zdahpxdrlfrz51nqy3zgcqicrmz0l-libXi-1.7.10-dev/include -I/nix/store/j97a7zpxj4fqni648zz369mdmr44sb8x-libXrender-0.9.10-dev/include -I/nix/store/wrqc8hc63z7h9h43m9bk6wys6m8vhr93-libxkbcommon-1.3.0-dev/include -I/nix/store/5r5yzwn6g1hms6x8lfbgw98k2q3ms4zf-xcb-util-0.4.0-dev/include -I/nix/store/g55p08rf2015pnv3vxags4x2lyfzdx6g-xcb-util-image-0.4.0-dev/include -I/nix/store/grszdh2mhv2vmxjs4cb6mcn7509w23r8-xcb-util-keysyms-0.4.0-dev/include -I/nix/store/zw64zbh837fncv6sch26a5clax4xnw9r-xcb-util-renderutil-0.3.9-dev/include -I/nix/store/4mx3p4laxfrn4a96hrmnqx7zi9yz002b-xcb-util-wm-0.4.1-dev/include -I/nix/store/dqjnf237plah6xbj3yibd4j199sz5acd-libGL-1.3.3-dev/include -I/nix/store/rh2q61k42s4jyf5bswidyxyj8wiqbfwx-qtwayland-5.15.2-dev/include -I/nix/store/9dlm6a8mnihxsid78vq8i75b7z2hwna0-qtdeclarative-5.15.2-dev/include -I/nix/store/v4q7h8l1y56vmappja5nm9a1hlbz7wiv-qtsvg-5.15.2-dev/include -I/nix/store/cxk8smlj2fzkf60yzkhwqs0s239l1dyj-chmlib-0.40a/include -I/nix/store/qdy5zcdxx836cn4iiwz19kz7i57pqnjr-hunspell-1.7.0-dev/include -I/nix/store/gdjgnjplvfzgffdsbi2xyzrxnz6lpnnf-hyphen-2.8.8/include -I/nix/store/phbm6a24xfz7fy6v499q6hbi2a77vsdr-imagemagick-7.1.0-2-dev/include -I/nix/store/35nxkhg1n97l00amz5vxw1b1nvf2907r-lcms2-2.12-dev/include -I/nix/store/g205b57giaak7zwjsv8qn9nb2xfjcy02-libtiff-4.3.0-dev/include -I/nix/store/0zjn0yz6hy9r57jscib0va2jyabv8v6y-xz-5.2.5-dev/include -I/nix/store/jjnfj8kiv47pqlnam00phamqmk3k5425-libXt-1.2.1-dev/include -I/nix/store/18nzsmi8pifmdkly7bh5d9497ls22589-libSM-1.2.3-dev/include -I/nix/store/sjy0n2jv8wmh3nyd1vp8spvkrwmbc1vr-libICE-1.0.10-dev/include -I/nix/store/sq5xwqxisd0favjpvr5crni7zwvarilg-libwebp-1.1.0/include -I/nix/store/anw18s6n6wxwhy9wlg91nz3ha6afvp29-libmtp-1.1.18-dev/include -I/nix/store/jmhq844mxk5qw07lh6a4myq86a417xzk-libusb-1.0.24-dev/include -I/nix/store/5b1xnpidzy10xr6in91nriiq2grbccmf-systemd-minimal-247.6-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/nwnxgqnw5ah2yxvxl9w3ql1mkfpl8j9x-poppler-utils-21.06.1-dev/include -I/nix/store/9g6jf45xylymikm64s1px9zrsa0l2bkw-openjpeg-2.4.0-dev/include -I/nix/store/c5vgx2wmnyx1a00jwpwyy7vad6yaqnhj-cairo-1.16.0-dev/include -I/nix/store/bz978lkvmigdb1diq4frnfr0dvihbzly-pixman-0.38.4/include -I/nix/store/ph7vn6prgils7vpa87dajxk7817gblsy-curl-7.76.1-dev/include -I/nix/store/d31bcv7pvwwy03cpnzv3sc0p3rhgxz29-nghttp2-1.43.0-dev/include -I/nix/store/g7aphzqms0p70bywwlnkhw9m9zs8iqvf-libidn-1.37-dev/include -I/nix/store/mhg4i7mpna2y0cx4m1yy0m6pybii5wmg-libkrb5-1.18-dev/include -I/nix/store/9rfg2ml9lmm7phlcyx8kzyc6m7sm0mpk-libssh2-1.9.0-dev/include -I/nix/store/519mf0va4bwvnppllmbg5m3izk4pz9l9-brotli-1.0.9-dev/include -I/nix/store/v1s51h6s92x7f2a614wgw2n0slg17gyb-nss-3.67-dev/include -I/nix/store/nv5jcki2q0zfk0xp5vd6n87sfrraiy5p-nspr-4.31-dev/include -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include -I/nix/store/y12vj44sgj7g53cvpalih92amwvcsxkc-python3.9-dbus-python-1.2.16-dev/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0 -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/x8664-unknown-linux-gnu -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/backward -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include-fixed -I/nix/store/bq095znl868pg40844laf50fv701s9k2-glibc-2.33-47-dev/include ../../../../../src/calibre/gui2/pictureflow/pictureflow.h -o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/pictureflow -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o mocpictureflow.o rm -f g++ -Wl,--version- -Wl,-O1 -Wl,-rpath,/nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib -shared -o sippictureflowcmodule.o sippictureflowPictureFlow.o sippictureflowFlowImages.o pictureflow.o mocpictureflow.o -L/nix/store/v9n1id9apyhc7midsxx1j45l6df922x5-libGL-1.3.3/lib /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- -L/nix/store/zf44435sdgyys6rxanbbyns62n7fzv7j-libglvnd-1.3.3/lib -lGL -lpthread cp -f pictureflow.cpython-39-x8664-linux- make[1]: Leaving directory '/build/calibre-5.17.0/build/pyqt/pictureflow/build/pictureflow' cd progressindicator/ && ( test -e Makefile /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -o Makefile /build/calibre- ) && make -f Makefile make[1]: Entering directory '/build/calibre-5.17.0/build/pyqt/progressindicator/build/progressindicator' g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sipprogressindicatorcmodule.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sipprogressindicatorQHash18000100QString.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sipprogressindicatorQProgressIndicator.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sipprogressindicatorSpinAnimator.o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o QProgressIndicator.o g++ -pipe -O2 -fno-exceptions -Wall -Wextra -dM -E -o mocpredefs.h /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2- /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/moc -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB --include /build/calibre-5.17.0/build/pyqt/progressindicator/build/progressindicator/mocpredefs.h -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -I/build/calibre-5.17.0/build/pyqt/progressindicator/build/progressindicator -I/build/calibre-5.17.0/build/pyqt/progressindicator/build/progressindicator -I/build/calibre-5.17.0/build/pyqt/progressindicator/build -I/build/calibre-5.17.0/src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include -I/nix/store/1yfjzqywgdrllq3cxvr5b7bkaa12ana4-libxml2-2.9.12-dev/include -I/nix/store/gxhfbfaqkh5pna8ap5xajjzlcj2dcp4y-zlib-1.2.11-dev/include -I/nix/store/wfwyqavhqwnqrjgq1jyhg2g9g6ya9awf-libxslt-1.1.34-dev/include -I/nix/store/qsl4jjnr5ji16nl636kdj24iiasj1ws4-openssl-1.1.1k-dev/include -I/nix/store/fiacxkszwy3gah3jx8c2g6rby5sj5g3c-sqlite-3.35.5-dev/include -I/nix/store/8z6xcnagxrgzs6hm5n7rvw7fqffgdlva-harfbuzz-2.8.1-dev/include -I/nix/store/6iwy5bxv8p64fizdd9m6277x8sbgaqq4-graphite2-1.3.14/include -I/nix/store/gn3qqpj4zi49pyjg5mf30aks07k49650-icu4c-69.1-dev/include -I/nix/store/qk9wf7rxzpi3b2yjsrglfy2ji482yscx-libjpeg-turbo-2.1.0-dev/include -I/nix/store/4kiqgcp4zjg25hpwm8mpv12hfi1qwiba-libpng-apng-1.6.37-dev/include -I/nix/store/63s3m6ifs7lcfx8icz7gz59qzabkww6z-pcre2-10.36-dev/include -I/nix/store/4ipg90ns4kx9p5qpwj0f6d6ysgf9x9lx-dbus-1.12.20-dev/include -I/nix/store/3lr9nqspakav45w936rb4y5q4hlps87b-expat-2.4.1-dev/include -I/nix/store/6l8vy0rpikcw0j1xw67qf983sqpxcfzp-glib-2.68.3-dev/include -I/nix/store/dxlyb8r1nxvjmak635rm1r654ig69z02-libffi-3.3-dev/include -I/nix/store/vyy5f2lnkq1b3r7rr79x1hwwcj2di3w6-gettext-0.21/include -I/nix/store/sxxkqm7smpdxwi985a047mvxdmgaw2bf-glibc-iconv-2.33/include -I/nix/store/hd68r4arz1smy4skafkvwzw5q2r61hk8-systemd-247.6-dev/include -I/nix/store/0fqvkh0lz98bn2zwiarfij48dzblg9r8-fontconfig-2.13.94-dev/include -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include -I/nix/store/lsfybn0mjjdx8qa6zf9n98xq2vj5fvy7-bzip2-1.0.6.0.2-dev/include -I/nix/store/s8dw25jfsqrx40q5m8rx8g4r48nnvbkb-libX11-1.7.2-dev/include -I/nix/store/z8k375ymlx56l6mhm6y37sfb8s06mxd0-xorgproto-2021.4/include -I/nix/store/mjcc6qjh7r729m3cnl8yh61pg07hqpxr-libxcb-1.14-dev/include -I/nix/store/brda65m3j4kym70mmvhn1j90gaysvpcs-libXcomposite-0.4.5-dev/include -I/nix/store/42ds8bnqg6sghqq6q4hm0f8xgryhk7q2-libXfixes-6.0.0-dev/include -I/nix/store/ibmnb203lx07sx840kqyjzjyms6kjplh-libXext-1.3.4-dev/include -I/nix/store/iyymggidpjja58ljzwl2vfbfyr8a8j4x-libXau-1.0.9-dev/include -I/nix/store/0b5zdahpxdrlfrz51nqy3zgcqicrmz0l-libXi-1.7.10-dev/include -I/nix/store/j97a7zpxj4fqni648zz369mdmr44sb8x-libXrender-0.9.10-dev/include -I/nix/store/wrqc8hc63z7h9h43m9bk6wys6m8vhr93-libxkbcommon-1.3.0-dev/include -I/nix/store/5r5yzwn6g1hms6x8lfbgw98k2q3ms4zf-xcb-util-0.4.0-dev/include -I/nix/store/g55p08rf2015pnv3vxags4x2lyfzdx6g-xcb-util-image-0.4.0-dev/include -I/nix/store/grszdh2mhv2vmxjs4cb6mcn7509w23r8-xcb-util-keysyms-0.4.0-dev/include -I/nix/store/zw64zbh837fncv6sch26a5clax4xnw9r-xcb-util-renderutil-0.3.9-dev/include -I/nix/store/4mx3p4laxfrn4a96hrmnqx7zi9yz002b-xcb-util-wm-0.4.1-dev/include -I/nix/store/dqjnf237plah6xbj3yibd4j199sz5acd-libGL-1.3.3-dev/include -I/nix/store/rh2q61k42s4jyf5bswidyxyj8wiqbfwx-qtwayland-5.15.2-dev/include -I/nix/store/9dlm6a8mnihxsid78vq8i75b7z2hwna0-qtdeclarative-5.15.2-dev/include -I/nix/store/v4q7h8l1y56vmappja5nm9a1hlbz7wiv-qtsvg-5.15.2-dev/include -I/nix/store/cxk8smlj2fzkf60yzkhwqs0s239l1dyj-chmlib-0.40a/include -I/nix/store/qdy5zcdxx836cn4iiwz19kz7i57pqnjr-hunspell-1.7.0-dev/include -I/nix/store/gdjgnjplvfzgffdsbi2xyzrxnz6lpnnf-hyphen-2.8.8/include -I/nix/store/phbm6a24xfz7fy6v499q6hbi2a77vsdr-imagemagick-7.1.0-2-dev/include -I/nix/store/35nxkhg1n97l00amz5vxw1b1nvf2907r-lcms2-2.12-dev/include -I/nix/store/g205b57giaak7zwjsv8qn9nb2xfjcy02-libtiff-4.3.0-dev/include -I/nix/store/0zjn0yz6hy9r57jscib0va2jyabv8v6y-xz-5.2.5-dev/include -I/nix/store/jjnfj8kiv47pqlnam00phamqmk3k5425-libXt-1.2.1-dev/include -I/nix/store/18nzsmi8pifmdkly7bh5d9497ls22589-libSM-1.2.3-dev/include -I/nix/store/sjy0n2jv8wmh3nyd1vp8spvkrwmbc1vr-libICE-1.0.10-dev/include -I/nix/store/sq5xwqxisd0favjpvr5crni7zwvarilg-libwebp-1.1.0/include -I/nix/store/anw18s6n6wxwhy9wlg91nz3ha6afvp29-libmtp-1.1.18-dev/include -I/nix/store/jmhq844mxk5qw07lh6a4myq86a417xzk-libusb-1.0.24-dev/include -I/nix/store/5b1xnpidzy10xr6in91nriiq2grbccmf-systemd-minimal-247.6-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/nwnxgqnw5ah2yxvxl9w3ql1mkfpl8j9x-poppler-utils-21.06.1-dev/include -I/nix/store/9g6jf45xylymikm64s1px9zrsa0l2bkw-openjpeg-2.4.0-dev/include -I/nix/store/c5vgx2wmnyx1a00jwpwyy7vad6yaqnhj-cairo-1.16.0-dev/include -I/nix/store/bz978lkvmigdb1diq4frnfr0dvihbzly-pixman-0.38.4/include -I/nix/store/ph7vn6prgils7vpa87dajxk7817gblsy-curl-7.76.1-dev/include -I/nix/store/d31bcv7pvwwy03cpnzv3sc0p3rhgxz29-nghttp2-1.43.0-dev/include -I/nix/store/g7aphzqms0p70bywwlnkhw9m9zs8iqvf-libidn-1.37-dev/include -I/nix/store/mhg4i7mpna2y0cx4m1yy0m6pybii5wmg-libkrb5-1.18-dev/include -I/nix/store/9rfg2ml9lmm7phlcyx8kzyc6m7sm0mpk-libssh2-1.9.0-dev/include -I/nix/store/519mf0va4bwvnppllmbg5m3izk4pz9l9-brotli-1.0.9-dev/include -I/nix/store/v1s51h6s92x7f2a614wgw2n0slg17gyb-nss-3.67-dev/include -I/nix/store/nv5jcki2q0zfk0xp5vd6n87sfrraiy5p-nspr-4.31-dev/include -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include -I/nix/store/y12vj44sgj7g53cvpalih92amwvcsxkc-python3.9-dbus-python-1.2.16-dev/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0 -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/x8664-unknown-linux-gnu -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/backward -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include-fixed -I/nix/store/bq095znl868pg40844laf50fv701s9k2-glibc-2.33-47-dev/include ../../../../../src/calibre/gui2/progressindicator/QProgressIndicator.h -o g++ -c -pipe -O2 -fno-exceptions -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNOEXCEPTIONS -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/gui2/progressindicator -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o mocQProgressIndicator.o rm -f g++ -Wl,--version- -Wl,-O1 -Wl,-rpath,/nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib -shared -o sipprogressindicatorcmodule.o sipprogressindicatorQHash18000100QString.o sipprogressindicatorQProgressIndicator.o sipprogressindicatorSpinAnimator.o QProgressIndicator.o mocQProgressIndicator.o -L/nix/store/v9n1id9apyhc7midsxx1j45l6df922x5-libGL-1.3.3/lib /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- -L/nix/store/zf44435sdgyys6rxanbbyns62n7fzv7j-libglvnd-1.3.3/lib -lGL -lpthread cp -f progressindicator.cpython-39-x8664-linux- make[1]: Leaving directory '/build/calibre-5.17.0/build/pyqt/progressindicator/build/progressindicator' cd imageops/ && ( test -e Makefile /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/qmake -o Makefile /build/calibre- ) && make -f Makefile make[1]: Entering directory '/build/calibre-5.17.0/build/pyqt/imageops/build/imageops' g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/utils/imageops -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o sipimageopscmodule.o g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/utils/imageops -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o imageops.o g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/utils/imageops -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o quantize.o g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DSIPPROTECTEDISPUBLIC -Dprotected=public -DQTNODEBUG -DQTPLUGIN -DQTWIDGETSLIB -DQTGUILIB -DQTCORELIB -I. -I. -I.. -I../../../../../src/calibre/utils/imageops -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include/python3.9 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtWidgets -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o ordereddither.o rm -f g++ -Wl,--version- -Wl,-O1 -Wl,-rpath,/nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib -shared -o sipimageopscmodule.o imageops.o quantize.o ordereddither.o -L/nix/store/v9n1id9apyhc7midsxx1j45l6df922x5-libGL-1.3.3/lib /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- -L/nix/store/zf44435sdgyys6rxanbbyns62n7fzv7j-libglvnd-1.3.3/lib -lGL -lpthread cp -f imageops.cpython-39-x8664-linux- make[1]: Leaving directory '/build/calibre-5.17.0/build/pyqt/imageops/build/imageops' Info: creating stash file /build/calibre- Project MESSAGE: This project is using private headers and will therefore be tied to this specific Qt module build version. Project MESSAGE: Running this project against other versions of the Qt modules may crash at any arbitrary point. Project MESSAGE: This is not a bug, but a result of using Qt internals. You have been warned! g++ -pipe -O2 -Wall -Wextra -dM -E -o mocpredefs.h /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2- g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DQTNODEBUG -DQTPLUGIN -DQTFONTDATABASESUPPORTLIB -DQTSERVICESUPPORTLIB -DQTEVENTDISPATCHERSUPPORTLIB -DQTGUILIB -DQTDBUSLIB -DQTCORELIB -I. -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtDBus -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o headlessintegration.o /build/calibre- g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DQTNODEBUG -DQTPLUGIN -DQTFONTDATABASESUPPORTLIB -DQTSERVICESUPPORTLIB -DQTEVENTDISPATCHERSUPPORTLIB -DQTGUILIB -DQTDBUSLIB -DQTCORELIB -I. -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtDBus -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o headlessbackingstore.o /build/calibre- g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DQTNODEBUG -DQTPLUGIN -DQTFONTDATABASESUPPORTLIB -DQTSERVICESUPPORTLIB -DQTEVENTDISPATCHERSUPPORTLIB -DQTGUILIB -DQTDBUSLIB -DQTCORELIB -I. -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtDBus -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o fontconfigdatabase.o /build/calibre- /nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/bin/moc -DQTNODEBUG -DQTPLUGIN -DQTFONTDATABASESUPPORTLIB -DQTSERVICESUPPORTLIB -DQTEVENTDISPATCHERSUPPORTLIB -DQTGUILIB -DQTDBUSLIB -DQTCORELIB --include /build/calibre-5.17.0/build/headless/mocpredefs.h -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -I/build/calibre-5.17.0/build/headless -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtDBus -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include -I/nix/store/1yfjzqywgdrllq3cxvr5b7bkaa12ana4-libxml2-2.9.12-dev/include -I/nix/store/gxhfbfaqkh5pna8ap5xajjzlcj2dcp4y-zlib-1.2.11-dev/include -I/nix/store/wfwyqavhqwnqrjgq1jyhg2g9g6ya9awf-libxslt-1.1.34-dev/include -I/nix/store/qsl4jjnr5ji16nl636kdj24iiasj1ws4-openssl-1.1.1k-dev/include -I/nix/store/fiacxkszwy3gah3jx8c2g6rby5sj5g3c-sqlite-3.35.5-dev/include -I/nix/store/8z6xcnagxrgzs6hm5n7rvw7fqffgdlva-harfbuzz-2.8.1-dev/include -I/nix/store/6iwy5bxv8p64fizdd9m6277x8sbgaqq4-graphite2-1.3.14/include -I/nix/store/gn3qqpj4zi49pyjg5mf30aks07k49650-icu4c-69.1-dev/include -I/nix/store/qk9wf7rxzpi3b2yjsrglfy2ji482yscx-libjpeg-turbo-2.1.0-dev/include -I/nix/store/4kiqgcp4zjg25hpwm8mpv12hfi1qwiba-libpng-apng-1.6.37-dev/include -I/nix/store/63s3m6ifs7lcfx8icz7gz59qzabkww6z-pcre2-10.36-dev/include -I/nix/store/4ipg90ns4kx9p5qpwj0f6d6ysgf9x9lx-dbus-1.12.20-dev/include -I/nix/store/3lr9nqspakav45w936rb4y5q4hlps87b-expat-2.4.1-dev/include -I/nix/store/6l8vy0rpikcw0j1xw67qf983sqpxcfzp-glib-2.68.3-dev/include -I/nix/store/dxlyb8r1nxvjmak635rm1r654ig69z02-libffi-3.3-dev/include -I/nix/store/vyy5f2lnkq1b3r7rr79x1hwwcj2di3w6-gettext-0.21/include -I/nix/store/sxxkqm7smpdxwi985a047mvxdmgaw2bf-glibc-iconv-2.33/include -I/nix/store/hd68r4arz1smy4skafkvwzw5q2r61hk8-systemd-247.6-dev/include -I/nix/store/0fqvkh0lz98bn2zwiarfij48dzblg9r8-fontconfig-2.13.94-dev/include -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include -I/nix/store/lsfybn0mjjdx8qa6zf9n98xq2vj5fvy7-bzip2-1.0.6.0.2-dev/include -I/nix/store/s8dw25jfsqrx40q5m8rx8g4r48nnvbkb-libX11-1.7.2-dev/include -I/nix/store/z8k375ymlx56l6mhm6y37sfb8s06mxd0-xorgproto-2021.4/include -I/nix/store/mjcc6qjh7r729m3cnl8yh61pg07hqpxr-libxcb-1.14-dev/include -I/nix/store/brda65m3j4kym70mmvhn1j90gaysvpcs-libXcomposite-0.4.5-dev/include -I/nix/store/42ds8bnqg6sghqq6q4hm0f8xgryhk7q2-libXfixes-6.0.0-dev/include -I/nix/store/ibmnb203lx07sx840kqyjzjyms6kjplh-libXext-1.3.4-dev/include -I/nix/store/iyymggidpjja58ljzwl2vfbfyr8a8j4x-libXau-1.0.9-dev/include -I/nix/store/0b5zdahpxdrlfrz51nqy3zgcqicrmz0l-libXi-1.7.10-dev/include -I/nix/store/j97a7zpxj4fqni648zz369mdmr44sb8x-libXrender-0.9.10-dev/include -I/nix/store/wrqc8hc63z7h9h43m9bk6wys6m8vhr93-libxkbcommon-1.3.0-dev/include -I/nix/store/5r5yzwn6g1hms6x8lfbgw98k2q3ms4zf-xcb-util-0.4.0-dev/include -I/nix/store/g55p08rf2015pnv3vxags4x2lyfzdx6g-xcb-util-image-0.4.0-dev/include -I/nix/store/grszdh2mhv2vmxjs4cb6mcn7509w23r8-xcb-util-keysyms-0.4.0-dev/include -I/nix/store/zw64zbh837fncv6sch26a5clax4xnw9r-xcb-util-renderutil-0.3.9-dev/include -I/nix/store/4mx3p4laxfrn4a96hrmnqx7zi9yz002b-xcb-util-wm-0.4.1-dev/include -I/nix/store/dqjnf237plah6xbj3yibd4j199sz5acd-libGL-1.3.3-dev/include -I/nix/store/rh2q61k42s4jyf5bswidyxyj8wiqbfwx-qtwayland-5.15.2-dev/include -I/nix/store/9dlm6a8mnihxsid78vq8i75b7z2hwna0-qtdeclarative-5.15.2-dev/include -I/nix/store/v4q7h8l1y56vmappja5nm9a1hlbz7wiv-qtsvg-5.15.2-dev/include -I/nix/store/cxk8smlj2fzkf60yzkhwqs0s239l1dyj-chmlib-0.40a/include -I/nix/store/qdy5zcdxx836cn4iiwz19kz7i57pqnjr-hunspell-1.7.0-dev/include -I/nix/store/gdjgnjplvfzgffdsbi2xyzrxnz6lpnnf-hyphen-2.8.8/include -I/nix/store/phbm6a24xfz7fy6v499q6hbi2a77vsdr-imagemagick-7.1.0-2-dev/include -I/nix/store/35nxkhg1n97l00amz5vxw1b1nvf2907r-lcms2-2.12-dev/include -I/nix/store/g205b57giaak7zwjsv8qn9nb2xfjcy02-libtiff-4.3.0-dev/include -I/nix/store/0zjn0yz6hy9r57jscib0va2jyabv8v6y-xz-5.2.5-dev/include -I/nix/store/jjnfj8kiv47pqlnam00phamqmk3k5425-libXt-1.2.1-dev/include -I/nix/store/18nzsmi8pifmdkly7bh5d9497ls22589-libSM-1.2.3-dev/include -I/nix/store/sjy0n2jv8wmh3nyd1vp8spvkrwmbc1vr-libICE-1.0.10-dev/include -I/nix/store/sq5xwqxisd0favjpvr5crni7zwvarilg-libwebp-1.1.0/include -I/nix/store/anw18s6n6wxwhy9wlg91nz3ha6afvp29-libmtp-1.1.18-dev/include -I/nix/store/jmhq844mxk5qw07lh6a4myq86a417xzk-libusb-1.0.24-dev/include -I/nix/store/5b1xnpidzy10xr6in91nriiq2grbccmf-systemd-minimal-247.6-dev/include -I/nix/store/bmafn4jpahgnzq30sfxkmhzga2axr2bs-podofo-0.9.7-dev/include -I/nix/store/nwnxgqnw5ah2yxvxl9w3ql1mkfpl8j9x-poppler-utils-21.06.1-dev/include -I/nix/store/9g6jf45xylymikm64s1px9zrsa0l2bkw-openjpeg-2.4.0-dev/include -I/nix/store/c5vgx2wmnyx1a00jwpwyy7vad6yaqnhj-cairo-1.16.0-dev/include -I/nix/store/bz978lkvmigdb1diq4frnfr0dvihbzly-pixman-0.38.4/include -I/nix/store/ph7vn6prgils7vpa87dajxk7817gblsy-curl-7.76.1-dev/include -I/nix/store/d31bcv7pvwwy03cpnzv3sc0p3rhgxz29-nghttp2-1.43.0-dev/include -I/nix/store/g7aphzqms0p70bywwlnkhw9m9zs8iqvf-libidn-1.37-dev/include -I/nix/store/mhg4i7mpna2y0cx4m1yy0m6pybii5wmg-libkrb5-1.18-dev/include -I/nix/store/9rfg2ml9lmm7phlcyx8kzyc6m7sm0mpk-libssh2-1.9.0-dev/include -I/nix/store/519mf0va4bwvnppllmbg5m3izk4pz9l9-brotli-1.0.9-dev/include -I/nix/store/v1s51h6s92x7f2a614wgw2n0slg17gyb-nss-3.67-dev/include -I/nix/store/nv5jcki2q0zfk0xp5vd6n87sfrraiy5p-nspr-4.31-dev/include -I/nix/store/yqjhkdarf7p74wcha7aypzsb751n7wl2-python3-3.9.5/include -I/nix/store/y12vj44sgj7g53cvpalih92amwvcsxkc-python3.9-dbus-python-1.2.16-dev/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0 -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/x8664-unknown-linux-gnu -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include/c++/10.3.0/backward -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/include -I/nix/store/lmwjr2fq2w875qn1in6wmzfirxkwjbnx-gcc-10.3.0/lib/gcc/x8664-unknown-linux-gnu/10.3.0/include-fixed -I/nix/store/bq095znl868pg40844laf50fv701s9k2-glibc-2.33-47-dev/include /build/calibre- -o g++ -c -pipe -O2 -Wall -Wextra -DREENTRANT -fPIC -DQTNODEBUG -DQTPLUGIN -DQTFONTDATABASESUPPORTLIB -DQTSERVICESUPPORTLIB -DQTEVENTDISPATCHERSUPPORTLIB -DQTGUILIB -DQTDBUSLIB -DQTCORELIB -I. -I/nix/store/ql88nzza9gbp98vyd1lsnm6f0qxvy4q1-freetype-2.10.4-dev/include/freetype2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtFontDatabaseSupport/5.15.2/QtFontDatabaseSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtServiceSupport/5.15.2/QtServiceSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtEventDispatcherSupport/5.15.2/QtEventDispatcherSupport -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui/5.15.2/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtGui -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2 -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore/5.15.2/QtCore -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtDBus -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/include/QtCore -I. -I/nix/store/zimb45pvmnj0p1pz5nhfsyxjb6850k71-libglvnd-1.3.3-dev/include -I/nix/store/6v8dklzhbphwwgkf1k4j4h89l5vr1ph5-qtbase-5.15.2-dev/mkspecs/linux-g++ -o main.o /build/calibre- rm -f g++ -Wl,-O1 -Wl,-rpath,/nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib -shared -o main.o headlessintegration.o headlessbackingstore.o fontconfigdatabase.o -L/nix/store/v9n1id9apyhc7midsxx1j45l6df922x5-libGL-1.3.3/lib /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib/libQt5FontDatabaseSupport.a -L/nix/store/ns3xjnh5sk8cbcw5ay4ynd9mb7i6qwmw-icu4c-69.1/lib -L/nix/store/zg4jsb0s6l9wnzlz8a20hwni9g67ynlc-zlib-1.2.11/lib -L/nix/store/hr033n2j5qyj9wdb48kwl7n1krmzyzw9-libjpeg-turbo-2.1.0/lib -L/nix/store/mfk76q6khi31iwylzj3dld5xza9jlc8q-harfbuzz-2.8.1/lib -L/nix/store/cp2smk79dih0vz099s470lz4ri0bybxv-openssl-1.1.1k/lib -L/nix/store/8cglsipdhl9mcrp5ldvwdwxzig2mh6sz-libX11-1.7.2/lib -L/nix/store/j82a9lvqn5nm5s06i92lyhgbqbrkyrg9-libXext-1.3.4/lib -L/nix/store/r8y8mw8s2ipfjmbxa5vf1jpfw7m2n6g0-libXrender-0.9.10/lib -L/nix/store/qpiaa91lh4dw6aiwhr155a8slv3s0irl-cups-2.3.3-lib/lib -L/nix/store/rsn05rq6mwcsk9qngbf4vcmxhgqghmhg-mariadb-connector-c-3.1.13/lib -L/nix/store/zf44435sdgyys6rxanbbyns62n7fzv7j-libglvnd-1.3.3/lib -L/nix/store/sxj8jgf7zx9m2m4q16gppac90phqbs2q-fontconfig-2.13.94-lib/lib -lfontconfig -L/nix/store/d7gldi24058s3jk0ssn4ghr7mj8w4nca-freetype-2.10.4/lib -lfreetype -lz /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib/libQt5ServiceSupport.a /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase-5.15.2/lib/libQt5EventDispatcherSupport.a -L/nix/store/i6z36ml0nm0vl93q9jn40bwjbi5b6r8g-glib-2.68.3/lib -lgthread-2.0 -lglib-2.0 /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- /nix/store/dgwn7s384vq2ar30lpnnmwz7k9l1rg55-qtbase- -lGL -lpthread mv -f /build/calibre- build took 135.0 seconds Running gui Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiling form /build/calibre- Compiled 87 forms Creating gui took 11.7 seconds Running install INSTALL paths: LIB: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/lib/calibre SHARE: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/share/calibre Installing code to /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/lib/calibre Installing resources to /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/share/calibre Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-device Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-meta Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-convert Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-polish Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/markdown-calibre Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/web2disk Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-server Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/lrf2lrs Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/lrs2lrf Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-debug Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibredb Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-parallel Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-customize Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-complete Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/fetch-ebook-metadata Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre-smtp Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/calibre Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/lrfviewer Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-viewer Installing binary: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/bin/ebook-edit Installing calibre environment module: /nix/store/j03yar7n727lmbhhfjrk2f4hf1bddgb6-calibre-5.17.0/lib/python3.9/site- Traceback (most recent call last): File \"/build/calibre-\", line 119, in (main()) File \"/build/calibre-\", line 104, in main (opts) File \"/build/calibre-\", line 203, in runall (self, opts) File \"/build/calibre-\", line 199, in runcmd (opts) File \"/build/calibre-\", line 154, in run self.runpostinstall() File \"/build/calibre-\", line 177, in runpostinstall from import PostInstall File \"/build/calibre-\", line 13, in <modulefrom import allinputformats File \"/build/calibre-\", line 18, in <modulefrom calibre.customize.builtins import plugins as builtinplugins File \"/build/calibre-\", line 752, in <modulefrom import SMARTDEVICEAPP File \"/build/calibre-\", line 2044, in <modulefrom zeroconf import (BadTypeInNameException, HASATOZ, ImportError: cannot import name 'HASATOZ' from 'zeroconf' (/nix/store/870agbnrkmv1p2wm1qb3abdbxi561pgd-python3.9-zeroconf-0.32.1/lib/python3.9/site-) </pre</details Maintainer information:\nCan confirm. Experiencing this aswell while updating my system.\nwill fix this"}
{"_id":"q-en-nixpkgs-064735d0af352ab1cb05bc3d651ce2eea78fc2e79f824d5cfe10edb2bee70a3e","text":", CVSSv3=5.3 (nixos-21.05, nixos-unstable) For Eclipse Jetty versions 9.4.37-9.4.42, 10.0.1-10.0.5 & 11.0.1-11.0.5, URIs can be crafted using some encoded characters to access the content of the WEB-INF directory and/or bypass some security constraints. This is a variation of the vulnerability reported in CVE-2021-/GHSA-v7ff-8wcx-gmc5. Scanned versions: nixos-21.05: ; nixos-unstable: . Cc"}
{"_id":"q-en-nixpkgs-39da40111134b3ba214de53fa999cc376177f0e83a4ff32d863a45bea75d26c4","text":"Shutter fails to build, looks like an issue with Imagemagick. Steps to reproduce the behavior: Expect that it should build, it was working on my last generation. If applicable, add screenshots to help explain your problem. Add any other context about the problem here. Maintainer information: Probably related to\n'shutter' is outdated in nixpkgs, perhaps we should start with a package update.\nThat would be nice but I think the problem is due to changes in perl packages\ni tested and fixes this\nI'll update shutter but will make the pr after the above pr has been merged"}
{"_id":"q-en-nixpkgs-e7c40286b36b63540af7a53dd9cf62ddb6693ea601613d8e1dd8de3bc57066ce","text":"Thought I'd try out Opentoonz because it looks cool, however I can't get the derivation for libtiff-4.0.3-opentoonz (patched version of libtiff for opentoonz) to build. Steps to reproduce the behavior: on NixOS 21.05 -4.0.3-opentoonz derivation fails to build to successfully evaluate and drop me to a shell prompt from whence I can run and try out Opentoonz. This is the relevant output from :\nthe build of opentoonz is also failing on master but for a different reason"}
{"_id":"q-en-nixpkgs-fe82821387a4cfc6a4bbc94bf2c12f250b7509c956091daec4473e9f62912d12","text":"tpm2-tss is unable to resolve the tabrmd loader located in the tpm2-abrmd package throwing this error when Steps to reproduce the behavior: the abrmd daemon and set the tcti environment interface to tabrmd tool or package using the tpm2-tss library It resolves the loader dynamically and works. I have narrowed down the cause to this commit making the dynamic loader only resolve locally from the library output. Reverting this in an overlay seems to have fixed the issue, but I am unsure if this would cause regressions as I am unsure why the dynamic loading got patched in the first place.\nI patched the tpm2-tss because when using tpm2 as a library, the modules would not be up in the dlopen load path which would break modules loading. A couple more details here:\nThis was a bit more tricky than I expected, while other python packages gets the dlopen library name patched with the exact path this is not possible in this case due to tpm2-tss doing the loading as far I understand. Unless I remember wrong I think it might be possible to dlopen with the exact path in python before calling tpm2-tss init and ld might be able to resolve the already library loaded library, but I might remember wrong. But this would also be undesirable since you cannot dynamically resolve loaders this way I assume.\nHa! So the issue would be that tabmd tries to look up libtss2-tcti- and this is provided by tpm2-abrmd. I completely missed that."}
{"_id":"q-en-nixpkgs-c27e9aec7df39482b683200b125bfcdc64e0085d57fbbe43f6cce66acf3ad8de","text":"When I try to open a file dialog in Bitwig Studio 4.1.0 I am getting that error: I can manually start from command line and I'm seeing the file dialog and it returns the selected file. Unsure if this is of any use but is the decompiled class that calls . Steps to reproduce the behavior: bitwig (or open any other file dialog in bitwig) GTK file dialog opens Maintainer information: EDIT: more information\nI am not a maintainer of bitwig sudio. Just because I changed some lines does not make me knowledgeable about this issue.\nSorry for that. I just did what the bug report template told me:\nI played a little bit around with that problem. It seems to be connected to a change of to . further references: As a quick workaround you can launch the application with: is using an embedded version of java 16.\n/cc\nThe cause may be here: We must not wrap :\nTwo ways to fix: Fix the embedded JDK. Replace the embedded JDK with package.\nThanks for having a look into this. I guess we should fix the embedded JDK as it's the way how it is shipped and it should be a compatible version. I tried to add to which does not improve the situation.\nI tried to do the same thing and I noticed that the exact error message changed to: but I can still run . Not sure what that means. EDIT: I've found in another package that they do run patchelf on jspawnhelper. I've also this to the postFixup: and that made it work. Not sure if this is the most elegant way to do it but I can open a PR and we can discuss that there. Thanks :)"}
{"_id":"q-en-nixpkgs-169071391b39f5ceeb9efb41928ba21bb2cd1087eeb635e179a6cd2072a77d97","text":"As explained also here: Steps to reproduce the behavior: Thunderbird 91 to install any extensions fail with \"Thunderbird has prevented this site from installing an unverified add-on\" Add on is instaalled successfully system: host os: multi-user?: sandbox: version: nixpkgs: Maintainer information:\nI can't find any obvious explanation for this, since both Firefox 91 and Thunderbird 78 using the same build expression can install add-ons fine. As a workaround you can use thunderbird-78, which is still present on nixos-unstable and seems to work fine.\nYes, downgrading seems to be working. However I would still have an up to date thunderbird. Also are you reproducing this on TB 91?\nI can reproduce the problem.\nAnd installed ones can't be enabled. (except for themes and dictionaries)\nHaven't tried my nixos instance yet but this is probably an upstream Thunderbird issue with linux - my test Ubuntu systems are all exhibiting the same problems with 91. Clean install or upgrade over 78 makes no difference. I can't even enable/disable themes. Predictably - and annoyingly - my Win10 Thunderbird 91 tests with addons all work fine.\nyeah Downloading extensions from in Thunderbird 91 works for me as well. Just not\nFrom the thread above, looks like: Thunderbird downloaded from works fine. Rebuilding with \"\" fixes add-ons and language packs verification issue.\nOk so why thunderbird does not trusts ?\nWell, it does if the thunderbird binary is from So probably some changes to how things are built.\nAfter rebuilding thunderbird with MOZREQUIRESIGNING= normal behaviour is only restored for new profiles I create - annoyingly my existing profile (the one I care about, obviously) is still unable to install or activate addons/themes. It's not a Mozilla issue as the tar ball of 91 direct from them works fine. It seems to be a common artefact introduced into the distro release by - that I know of so far - NixOS, Slackware, Fedora and Ubuntu. Which is pretty weird.\nHere's the upstream bug report: In the mean time, distribution build changes are the only way to fix this.\nCannot, have never been able to add those tempting add-ons. Thunderbird says 'this add-on will have access to , etc. but clicking never does anything"}
{"_id":"q-en-nixpkgs-bacc995c1766244bb529e47fac43352c1c55ae3657a7da74b1c23a569313bcae","text":"Boto seems to be broken at least when brought in via the google-compute-image package. Steps to reproduce the behavior: the package in a derivation. to build the derivation. the error: I expect the build of the derivation to succeed. output forthe error: I'm not sure who the maintainers are. Please run and paste the result.\nThere's a PR to fix it by specifying a different Python version:"}
{"_id":"q-en-nixpkgs-93c6f704bc4d3029d9d3b2e362a1a117a32fc126fdd69e0a1ae58d2c779b33c5","text":"the nm-applet from current channel-nixos errors out with (.nm-applet-wrapped:4399): GLib-GIO-ERROR **: Settings schema '' is not installed zsh: trace trap nm-applet there is a open pull request which also wants to fix schemas. might this address the same problem? cheers, jonas\nUnfortunately the schemas problem needs to be addressed per-package (we can't have one global directory containing all schemas in the system). I don't understand this case, as everything seems fine. The schemas are in , they are compiled, and is wrapped to have them in search path.\nSame issue here - I just rolled back for the moment.\nthe culprit is pull request . strace shows that only compiled schemas are consulted and the pull request deleted those in preFixup (under the surmise of it being caches). pull request including a partial revert follows shortly."}
{"_id":"q-en-nixpkgs-25b9036f0643dfe2f50b6e7fb80307702aafcab9439b9d7c1de36031b93b9584","text":"cannot access (read/write) on first start. Steps to reproduce the behavior: systemd service Maintainer information:\nAlso reproducible on master. Looking into it.\nThank you :-)\ncontains a potential fix :)\nWorks perfectly! :+1:"}
{"_id":"q-en-nixpkgs-7aab85b7b0902068440b382b600b0fcc78518ded77a2670f501055b92a4b2c8d","text":"When the option is toggled, Tweaks log the following error on startup: GNOME Tweaks General tab, toggle Suspend when lid is closed It should prevent laptop suspension. The binary needs a python wrapper. Additionally, the setting creates which contains a store path, which might break on garbage collection. cc\nFor me the error is because I have a file which causes the module to exist but not have anything in it. Doesn't it have one? My take was that needs a dependency on the python package.\nEven if you have python modules linked to your system profile, Python still should not pick them up. Unless something is contaminating the environment variable. By default, only executables in package’s directory are wrapped:\nOh yeah, I forgot. It's me. I'm the one contaminating the . I did it on purpose a while ago and it didn't occur to me that that's why my exception was different than yours, or I wouldn't have left my previous comment in the first place. Oh, I see. Thanks.\nPlease test\nin the built package this is in\nI opened for this:"}
{"_id":"q-en-nixpkgs-7156ab47d5835df817bba15c06c28cef42ad0f7083f838709c4c8578f22d094b","text":"The gitlab service does rotate its log files which has potential of filling up the partition. Steps to reproduce the behavior: something that causes a lot of logs to be written. the partition runs full. Rotation of the logs similar to <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Maintainer information:"}
{"_id":"q-en-nixpkgs-8d9990a80cf372d63fc0fd8f3d16e9315bcf45e4f4e7cca1c2589e90f6ffb22b","text":"hangs when given a derivation because of infinite recursion either error or treat it as a string, which all other generators seem to do <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Maintainer information:\nis there a context where a derivation would even be a valid input for this function? If not perhaps a simple fix would be to test for a derivation as input and simply throw a warning instead of continuing computation.\nit shouldn't be possible to use derivation as a attrset for this function, but store paths could be useful which is similar to 's behavior\nSo this worked without error: I was able to reproduce the issue by just passing your example input above though, and in looks like nesting the derivation deeper also still causes the error. I guess we could map over the input at the start of the function and convert any unconverted derivations to strings before moving on. That should allow us to declare store paths and avoid this issue. Side note: you shouldn't have to restart your system to recover from the error condition; worked for me\nI forgot SIGKILL existed. I removed the warning in the description\nI have run into a very strange issue. I thought that the following patch should be all that is required to fix this: however even though I confirmed in repl that the conditional evaluates to false when passed `\"${hello}\", the codepath guarded by the conditional still seems to run anyway and fails with the following error: to add to the confusion, if I push the value one level deeper then it works fine:\ni don't think ini support values at top level, or maybe its just the ini generator?\nah okay, then I guess my change is completely ineffective. I'll keep iterating :laughing:"}
{"_id":"q-en-nixpkgs-e5ca665bfefea16ae352dda3b965f1f7159d15c981848046f0a6d5acd967e67a","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on github to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- [x] I am a noob and tried to fulfill the above requirements to the best of my abilities and understanding of github name: pyradio <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 0.8.7.2 desired version: 0.8.9.9 at least 0.8.8.5 (last of 0.8.8 chain) <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR. The current version (0.8.7.2) is crashing for me when increasing/decreasing the volume. (Fixed in 0.8.8) The newest 0.8.9 versions feature a radio-browser-overlay, which would be very nice.\nPR in"}
{"_id":"q-en-nixpkgs-1bd2982d0b48ca7f82359452ceadcd237e57bebf1bcee048dce7e24b5fa8db03","text":"Project description A terminal UI utility for using systemctl interactively via fzf. Metadata homepage URL: source URL: license: Unlicense platforms: linux, darwin\nSubmitted a PR"}
{"_id":"q-en-nixpkgs-e72ed13735ba03a0be4aa7b4271cacb3b9a3ee51a031a54268a1fdfd6c1f6dd2","text":"cc\nBisected to\nThis is an error when calling . We should fix git wrapper."}
{"_id":"q-en-nixpkgs-2ad694f6e9e0ec49bcef78098ac262208c4a25338a7d4c0c3a6e60abc7fa6652","text":"It doesn't work with nor with a flake. Works fine on my NixOS machine. I don't even know what means. With a flake I suspect that this is related to . I've had so many dynamic linking errors in the past week, especially from Rust stuff, that I no longer have the energy to look into this myself, hence the bug report. Please run and paste the result.\nCould you (or anyone with access to aarch64-darwin) try another run against my branch?\n$ nix-shell -p zola error: missing bootstrap url for platform aarch64-apple-darwin (use '--show-trace' to show detailed location information) is coming from and since the flake works you should update your channels\nI'm trying your commit but it might take a while for things to compile. I'll probably ask about this on Discourse since it's unrelated to this issue but I was hoping that my Darwin machine would run the I install through the flake that makes up my dotfiles. So hopefully taking channels out of the equation, but I think that's not how it works, so I'll look into this, thanks.\nthat should be possible but uses channels while uses flakes. remove your channels to be sure that you're using flakes instead of channels\nworks, thanks! If someone needs this fix right now\nThanks Here's another way of applying the fix that doesn't require a overlay, which works for my use case:"}
{"_id":"q-en-nixpkgs-c9d555bb97da1317575c6235ad1e6883a1b96aa934a44da1521408ed8e32570e","text":"Project description It seems that language packs can't be installed using . There are 14 languages in total available. I've tried to pack them myself but I am really new to NixOS and failed :) Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin\nI have these on my fork and will cleanup/submit a PR soon.\n(or someone with write permissions), you can close this issue. PR has been merged, but wasn't linked to this issue until after the merge so it didn't auto-close the issue."}
{"_id":"q-en-nixpkgs-7099b23863a014eb7ec299a4c6597718eedb05e129bfcd1d2b33ced726c76ba1","text":"In , the patch was removed. Without this patch, trying to restart qtile just kills it and goes to the display manager, closing all of your apps. Steps to reproduce the behavior: into qtile any key combination to to press this combination Qtile restarts without closing any apps Qtile just quits and goes to the DM Maintainer information:\nThis patch can fix the issue\nI think most of this can be done in the wrapper. I was hoping that buildenv would solve the issue as the python interpreter would have all of the dependencies it would need.\nI dug a little deeper and found the cause to be the following: Qtile restarts with sys.executable, which leads to the following full command on my machine (ignore the package name, not important here): This results in a syntax error So it's trying to execute a bash script through python, which leads to a syntax error. I see some related issues on nixpkgs, , Do you have any idea how to continue? Edit: To add, we could re-introduce a patch like this (this is for current qtile master) But the whole idea was that we wanted to get rid of the patches :/\nI'm okay with doing minor patches like that. It was just before were were modifying the entry point significantly. But I'm also not a qtile user, was just trying to get qtile to a more recent version. If you want to create a PR, I'll be happy to review it."}
{"_id":"q-en-nixpkgs-f583accec0c5312c53b6a682772ccc1a4c80257b70c329b98bceb27a91809c88","text":"To make sure that we are on the same page: Kernel: (run ): System: (NixOS: , Ubuntu/Fedora: , ...): Nix version: (run ): Nixpkgs version: (run ): fails with Build succeeds Build fails\nI have the same issue."}
{"_id":"q-en-nixpkgs-20704114faf5304901434327b21ae9eb42f67e92e5fe88dd00b9257420349ddd","text":"I set up an box and want to set using , but I get an error: . outputs nothing. outputs: If is enabled, will be enabled by default. It will be great to enable for the default kernel so that people don't need to build their own kernel to use . Maybe this is related. Steps to reproduce the behavior: up an box to enable using get an error no error appears and is enabled My related config is as follows: no maintainer for Maybe is interested. Maintainer information:\nHah - I am definitely interested in this - AFAICT this keeps alternating - i.e. I've seen that sometimes the kernel enables it and sometimes it doesn't - I think the critical bit is actually to add a test to nixpkgs to ensure this is enabled, then fix the kernel config.\nI also set up an box with the same config and find that I can set bbr. It's wired. I tried to nixpkgs but didn't find out what cause the difference.\nIt turns out that the default kernel config generated by on does not enable . And on , is enabled by default: .\nI think your still has great value. Would you like to reopen it?"}
{"_id":"q-en-nixpkgs-f2eeed6b540912d4aa73fbab5bae50d4b530ad7310a9c383777bb5c1d7237047","text":"Project description A minimal Matrix chat client, focused on performance, offline functionality, and broad browser support. Metadata homepage URL: source URL: license: Apache platforms: Browser Deer matrix-team maintainers:\nI marked this as stale due to inactivity. &rarr;\nNo reaction, but still interresting for me.\nftfy:"}
{"_id":"q-en-nixpkgs-5e401a3aa68f0433fdacb00c2e92dfc21c7ffce0f84b4d040c2310008e87debc","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on github to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: whitesur-gtk-theme <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: \"2021-07-27\" desired version: 2021-09-24 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-908b3eedbbad67f28241ccb54998f54ea2d17606bbe5298f37676a0597ce3414","text":"Neovide fails to install via Steps to reproduce the behavior: up a terminal (currently I'm on aarch64-darwin, but I think this should be reproducible on different platforms) (or use home-manager/nix-darwin/nixOS) for it to give the mentioned error Neovide should install and run as expected <img width=\"1398\" alt=\"image\" src=\"https://user-\"(text form) Might be a result of this: Please run and paste the result. Maintainer information:\nCan you try this?"}
{"_id":"q-en-nixpkgs-70c680be64ef34e64a07debe5af7dce5064c5995e7b33715dd945f9e4405aeca","text":"Whenever I open a file chooser dialog in Nicotine+, it crashes with this error: Steps to reproduce the behavior: X environment without a DE and run a file dialog I saw a similar issue on Chromium: Maintainers (thanks!): system: host os: multi-user?: sandbox: version: channels(root): channels(ao): nixpkgs:\nIt looks like the glib setup hook isn't setting for some reason (maybe it's not running?) The wrapped file looks something like () So the packages is being to , which happens here: Right above it, gets to if it's present, but that doesn't show in the wrapper, so the variable is not getting set."}
{"_id":"q-en-nixpkgs-75c1d393b74b29f186ac9b604f809b10b3071452308074024f0e7f8b5f4f0133","text":"When I update nix I can't copy to our binay cache anymore. All I get is I bisected nixpkgs and found that the bug was introduced by . Run The parent commit works: This error here is fine because our cache has a different name of course. Please run and paste the result.\nJust wondered if here is the right place for the issue but I’m quite lost on this. Maybe this is a nix bug?\nNo you're correct to open this here, this is a good place to start. I'll try to bisect further to see which aws-sdk-cpp commit broke this functionality, hopefully that will tell us more about how to proceed.\nBisected to , but unfortunately the commit is very large so I don't see an obvious cause. Maybe it's best to take this upstream to the Nix repo now.\nThanks for reporting, and sorry if this broke your workflow. I've bookmarked this issue and I'm going to look into adding an installCheck regression test to ensure s3 stays functional on future updates.\nThank you all for the quick response. This was a really sneaky one. Luckily nix allows rollbacks. That would be great. I’m sorry I cannot think of anything better than checking the error message of . When I do I let you know."}
{"_id":"q-en-nixpkgs-bb1401208d6cc44f8fe318ae4301eed4692697f86a9d6beae4f9026b2ad97f88","text":"Usage of flag without , and it is known in the Clojure community that its usage causes issues. This is because without the binaries are statically linked with , and I tried to create a GraalVM derivation compatible with . The result is here: However, this can't be upstreamed since it brokes all compilations with (including dynamically linked programs as used in nixpkgs). The major issue is that, to make compilation work, . This approach works fine for when we have only one lib C, however if we have two separate lib C we will have name conflicts and either we will have a compilation with glibc symbols on musl libraries or vice-versa. So, what are the options? One approach I can think is extend the wrapper from binary to conditionally include either or depending on the context (e.g.: if the flag is passed or not). We can do this using the flag, as below: For this to work we also need to remove the symbolic links to glibc library, otherwise this can conflict with musl libraries. This should also fix this . However, this looks like a very ugly solution, and I am open for better ones. Steps to reproduce the behavior: . Keep in mind after , it expects binary as . You can create a symbolic link and add it to your a file with contents: (segfaults) It should compile and run successfully. N/A Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Maintainer information:"}
{"_id":"q-en-nixpkgs-ed5b072c80035be0d424c8e3cef69220465110b27db22da0a968966a4ec8841d","text":"I updated today my distribution on latest unstable (I was already on unstable before). And it turned out that many things suddently broke (the clock is broken, my does not work anymore (of course), Carla's software does not have anymore a \"Canvas\" menu, NSM can't run Carla or it gets an error, Giada also fails to connect to Jack...). I quickly realized that I was running Wayland instead of X11, while I was always running X11 before on KDE Plasma. Moreover, the KDE clock and calendar is now broken (negative numbers+hour does not appear at all+what is this ): ! So here are my questions: it normal that Wayland is the default now? I can't find any documentation on Wayland vs X11, and it seems a bit risky to use Wayland by default given all the bugs I encountered... yes, how can I configure X11 to be the default? (I mean in the configuration file, not by just changing at connection time) can I get the clock and calendar back? idea why Carla, Giada and NSM are also not working properly with Wayland? (I'm using pipewire) Steps to reproduce the behavior: the system and reboot Please run and paste the result.\nSame issue here. On top of that, I'm unable to share my screen with tools such as Google Meet, Teams, Jitsi. To fix: , then in SDDM select on the bottom left:\nThere's not supposed to be any changes to the default. If you don't set one via , you might get the wayland version. This should/must be fixed. Others have mentioned the issue with the clock although I'm not seeing that here. As for the screen sharing question, applications need wayland support. That really has nothing to do with Plasma or NixOS.\nThanks for your answer. How can I find a list of valid ? I tried to put an absurd value but it does not give any error. In my case I'm running sddm, plasma5 and x11. Concerning the clock, I'm not sure what I've that you don't. If it helps, the language of the system is french . and in X11 my clock looks like: ! The is replaced with octobre, so that's why I'm thinking it may be linked with the language.\nOr well, pipewire support. However screensharing/recording on plasma wayland seems to be broken at the moment (even in OBS which does have explicit support for wayland). Judging from what gets logged in the journal it seems like is not agreeing with kwin about the supported wayland protocols or something.\nSame, I also get the broken clock, unless I remove the timezone setting from and get UTC. I saw some people saying Plasma's systemd startup fixes it? Altho IIRC the PR for that was merged but the systemd service isn't running by default.\nI have which should be the most standard one... and.\nMaybe NixOs does not like people with European timezone…\nI notice that the env var is somehow missing for . I run with correctly set, and the clock widget just works. Running it with empty reproduces the same broken clock. I checked that , have , but , and don't. Since is the session entrance started by (so it's not 's fault), we can know the envvar somehow gets dropped during the plasma starting chain. EDIT: Another issue is that, we used to be able to workaround the clock issue by . But after , as I mentioned in , and the option DO NOT work anymore. Session is now always booted without systemd.\nBy comparing between , exactly is dropped when the wrapper running , no other environment variable are , removed or modified. is wrapped to gain CAPSYSNICE, I'm not sure how the work, but simply removing this wrapper just fix the clock widget issue. But it's definitely not a good idea since is expecting to be real time scheduled. cc\nI have nothing to do with KDE, but I made a few changes to in . I don't see how this could possibly have affected the clock, but you could try reverting my commits if you think this is related.\nThat may be unrelated. I found the related upstream issue The cause here is that rejects , and some other env vars in secure-excution mode, including programs with or capabilities. See In our case, has capabilities set. But I'm still not sure how we could workaround this. Currently my thought is to patch and to let the wrapper pass these envvars in other ways, like renaming envvar name or passing through arguments. If there's a better way to workaround, please point it out.\nThe \"Wayland by default\" thing was introduced here: as part of KDE Wayland support in NixOS in general. Mb shouldn't be default yet or at least not to override previous preference of X11 during upgrade. +1 about the clock - same issue. I'm on locale.\nCurrent status: The clock issue can be workaround by setting since which is already in . It will be fixed (without ) in The default session will be set back to X11 on SDDM in Currently, you can manually switch to X11 and login once, and SDDM will remember your last choice (hopefully, it may be reset if you change available desktop sessions. It's an issue of SDDM and I'll look into it later).\nThis fix didn't work for me. The clock is still broken in exactly the same way as reported in this bug. Even after logging into fresh user. (Again, I'm on Wayland)\nAre you sure on nixpkgs and that option is enabled? Please check if the plasma systemd service is started by . If it's not running, please open another issue for not working. If it's running, then there may be some issue of . Since will load PAM environment, which must contain , and it will be propagated to other units. In this case please open an issue about \"systemd user units not picking PAM environment\".\nThanks for your advice! Yes indeed I'm on master (last synced yesterday) and have this option enabled: Though it says \"dead\" for some reason (and I'm writing this from an active KDE session), process is a child of systemd according to pstree TZDIR is defined as well: Do you have any other ideas?\nLooking at 's process env, indeed env is not defined there. Is that the issue you are trying to point out? (The above returns nothing)\nPlease open an issue about not working, and provide there.\nDid another update from unstable (it updated LOTs of packages) and after reboot, the clock is working again! (and TZDIR env is indeed known to plasma). thanks again for jumping on this!\nCould this be the same root cause for issues with IMEs, screen capturing and Flameshot simply not working at all? IME (fcitx) does not activate, screen capturing in browsers and OBS does not show a source selection dialog, Flameshot does not react to its DBus invocation (Spectacle works although it is dog slow -- not sure what their internal differences are) See and I don't know much about DBus but the bit about it not finding services dirs sounds like it could be caused by something like this. I also remember seeing some logs from Flameshot where it thinks it has been started already but actually invoking its screenshot function does not do anything -- wrong DBus instance? ... Or am I the only one with these problems on plasma/wayland? :D\nI have 4k 14\" laptop display and 2k 27\" external display. Flameshot doesn't work for me at all on KDE+Wayland: Spectacle kind of works - about every second time; plus, when trying to annotate, the resolution of the image gets absolutely trashed if I the screenshot was taken from non-hidpi screen.\nFor the issue, see\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-ab84ecf27959c0deeb7a10f692b450dfdec0e3be296645133b478bdc4910caeb","text":"It seems that all of a sudden a few dependencies are missing to . As a start, I found the need to add . The problem is with which also makes the impression it's needed: But our build of doesn't provide cmake files. Building with cmake instead of autoconf might help, but they don't use the cmake module there which makes the transition harder. cc (maintainers) and .\nThe new requirements come from thrift since\nI have an idea of how to fix this. Will put a PR up shortly"}
{"_id":"q-en-nixpkgs-9cf14a22dbe8eeb899dbd45adbc57e860093f1c25910b791b334431ae80c808e","text":"Except from If the main module specifies go 1.17 or higher, go mod vendor now omits and files for vendored dependencies, which can otherwise interfere with the ability of the go command to identify the correct module root when invoked within the vendor tree. This may change the hashes of vendorSha256 in existing packages that are using buildGoModule. Originally posted by in\ngo 1.17 may not be portable across different go versions, but go 1.16 will be EOL next month (when 1.18 releases). 15 is EOL and should be removed. There isn't a remediation for this problem correct? The only answer I see is to move the versions forward with breaking changes.\nYeah - Regenning the hashes isn't that hard - I'm going to confirm they're stable, and then stack a PR using the infra I used for the og hashes. Walking through the cases vendorSha256 = null Fine - no vendoring happens Go 1.16 to 1.17, not go1.17 No module change - testcase clipman. go 1.16 to go1.17, is go1.17\nI'm already working on this.\n- Do you have an ETA? I expect I can get this pounded out this weekend, but if you don't want me to I won't.\nAlso any objection to me trying to clean out the go1.15 deps? I think some of them (i.e. terraform / prometheus), might not be updateable, but the rest probably are.\nHopefully in the coming week. We need to handle the hashes and the usual breakage from the update so I really don't want to push something though quickly in case it turns into a mess. I'm also avoiding normalizing any of the vendor hashes so we aren't needlessly adding noise to the git history. I've been moving some things to 1.16 or 1.17 in advance to limit diff and rebuilds and checking that the updated hashes still build on darwin+linux. If you don't mind doing the hashes in smaller batches like in that would be helpful? Fixing or pinning any broken packages would also help. No that should be fine, only looks like a few packages plus the pinned which needs to be either unpinned or removed.\nI'm happy to cut small PRs - How many hashes do you want them batched to? 10 hash updates per PR, 100? I have no strong opinion on how to batch these (and normally my tooling generates a commit per package, so its' real easy for me to batch arbitrarily). I'll probably try and clean out most of the easy stuff (i.e. everything with a null vendorSha, everything which doesn't have a changed vendorSha), then deal with the stuff which causes mod vendor changes, but if you want I can start on the gnarlier stuff first (modules with vendorSha changes + the non module stuff).\n20 or less per PR would make merges simple as we can let test it. Unless the needs to be updated or it's broken on 1.17 I'm leaving these for the final PR when the default is changed. Also anything that causes more than a few rebuilds is probably easier left for the final PR as well. I'm not planning to fixing every first before the final PR when the default gets changed but if we can get it down to 50-100 hashes it'll make it easier to test and also avoid conflicts with other PRs.\nGot it - so focus on the changed hashes - or anything which breaks. I also just pushed out working PRs for every go1.15 package except one (mostly to go1.17), so we might be able to get rid of go1.15 before go1.18 lands.\nShould be under 100 now but I'll keep going until there is only a few left (basically just stuff that should stay in sync with the default or causes a few rebuilds) as I've found some with broken vendoring on 1.17 and it's easier to address that first so the final PR is fairly simple."}
{"_id":"q-en-nixpkgs-9ddbdd4b83a601ebfbc27b8598a63282992307ce8d25104e84db614d991932c9","text":"Any local users can talk to locally running X server by default. That includes sending keyboard/mouse events, popping windows, etc. Log in as into a host running an X server under , and run I think normally you should have to do something like before you're able to talk to 's X server. System: NixOS: nix-shell -p xorg.xmessage --command xmessage test Nix version: nix-env (Nix) 1.11.2 Nixpkgs version:\nThanks, amazing nobody has spotted this before. The flag has been present ever since (when it at least had a comment about being bad idea...).\n:+1:"}
{"_id":"q-en-nixpkgs-47a25c0b9e15ed042e7c88cdb362a2ce5345eadf2116f5381664abae82a59ef4","text":"YouTube tabs reliably crash within a few minutes on nixpkgs-built firefox Steps to reproduce the behavior: firefox from latest (NixOS/nixpkgs to some youtube user's video list up and down for a bit This doesn't happen with my previous env (using firefox 92.0.1, I think NixOS/nixpkgs but I haven't confirmed), nor with . It does happen with . Also, the page is not available on the nixpkgs-built firefox. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nI think it happens when you hover over a video and it starts showing the preview.\nYep that does it!\nPossibly this bug?\nSeems familiar. Will try the fix they suggest.\nNote that in bisecting I've found a commit with the same version of ff but not this bug.\nI don't think it's the same bug, I'm able to watch both linked videos (as long as I don't mouse-over a preview on the side)\nIt's happening for me as well, not sure why though\nNot happening with firefox-bin 94.0\nAs said in the issue it doesn't happen with firefox-bin at all (also checked that myself) (also not 94.0.1)\nAlso happening for me (using nixpkgs.firefox).\nHotlinking directly to webp files also recreates the issue. That should narrow it down a bit. [edit: I guess these links expire after some time, but its just the url to an animated youtube thumbnail]\nAlso works on firefox-bin 94.0.1 (master is on 94.0).\nSeeing crash locally as well on and . I bisected down to (had to apply to avoid build failures). It is a merge commit with the following conflict resolution:\nWhile at it how does one get meaningful debug symbols out of firefox? is very sad: I tried patching in the below, but got no symbols: Is there a generic way to get the symbols? Or maybe it's a good opportunity to add them to specifically?\nI tried reverting , still broken. Worked when 94.0.1 was merged in .\nYeah, I have no idea on how to accomplish that either. And having that could be really helpful :+1:\nI can confirm this occurs when mousing over YouTube video link/thumbnail pairs in either the main page, search results or video sidebar (on Firefox 94.0.1). Not only that but doesn't exist at all, which probably means the crash reporting feature is entirely disabled for nixpkgs Firefox?\nWith the stdenv patch, does ff work on both parents of the merge commit but not the commit itself?\nwe have crashreporter turned off by default for some reason\nits been that way since (2008) we should set it to\nCrash report: https://crash-\ncorrect. Re-tested again just in case: ^0: - crash ^1: - ok ^2: - ok ^3: <invalid, only 2 parentsNote that these two branches contain about 2 weeks of commits. One of them is a update, another is an update. This merge mixes both together and firefox does not seem to work. is a full picture of all the commits involved. Anecdotally I tried to disable lto with and got working (but I suspect it also flips to in that case). Would be nice if controlled just aspect and not the rest of toolchain.\nAlthough using is possible for LTO ultimately it complicates the expression and deviates from upstream. When I implemented support for LTO, I think it was agreed upon that we should try to follow upstream as close as possible. Hence enforcing and . The reduced maintenance burden of having to test more configurations was also a plus. That being said it would be nice if someone wants to take on that burden and expand the LTO support to be more general (and maintain it). The end goal was to actually change to build to using too but allow the native system linker. I never got around to implementing that. The reasoning being upstream tests vigorously against . This is speculation on my part but I think a point of interest here is Firefox now builds against instead of . Building with was never a requirement for LTO but a side effect of the old stdenv. At the time did not have a proper nixpkgs wrapper script and was unable to be used outside of it. (On a side note, for some reason had issues during linking with not being configured to pull in LLVM's hence the workaround. It can be removed now since is being used.) reports (about:buildconfig) being built with LLVM 12 and rustc 1.55 so this might be a new bug introduced with LLVM 13 or the newer rustc. A definitive way to test if it is a bug in the switch to would be keeping the same firefox version and having applied (so is used) but change the underlying LLVM and rustc versions to 12 and 1.55 respectively. If the crash goes away then it is tied to the LLVM and rustc version if not then something wrong is happening when using .\nI created a upstream bug 3 days ago, including stack backtrace with symbols. The crash function is in . Disabling webp () can mitigate but it of course quite hurt. :disappointed: If anyone gets more information, please also post in the bug thread. BTW: Enabling debug symbols need to set configure flags and set . won't work due to .\nIt is mentioned in that bug report that directly opening links canot reproduce the crash. I actually had the opposite experience. I was able to recreate the crash by directly opening a webp file, although it's not 100% consistent. I also stumbled across this other URL that crashes firefox, this one is a PNG. If I right-click this URL and \"Open in new window\" the tab will crash about 50% of the time. Curiously, this URL seems unaffected, despite the resulting file being hash-identical. !\nI just now noticed that the above image is being served with a content-type, so looks like its still webp-specific, even though no webp content is being actually rendered.\nI cannot reproduce with that URL at all after tens of tries. This DOES crash me sometimes. And note that the URL actually responds a WEBP image despite it contains , maybe due to HTTP fields. You can confirm it by the tab title (once you don't crash) or download and check with . This is also fine for me. It's more complex than I think. I'll guess it might be an index-out-of-bound or other kind of heap corruption? Could you post about this in my bugzilla report?\nCan you share exact change needed? Here is my non-working example of enabling the symbols:\nYeah, whatever optimization tricks is doing here is just confusing things. I'm going to ignore those results for now. I was able to create a very small poc for the bug that works reliably with the same image, though. (Also posted in the bugzilla thread)\nFor how to build with debug symbols, see I'm still working on this issue now and I'll make a PR about that later. Note that you need to follow and set environment before starting or .\nThis is happening for me.\nSee\nthe in the first post is a typo, right? If I read it right, we can keep stable nixpkgs as it is, thanks to using LLVM 12 for firefox.\nYes, updated."}
{"_id":"q-en-nixpkgs-0316cf765ee2469f25fe4f0a5e5f097e7b4b9b68f0c8c2aba9908b4ac9d26f9a","text":"The runuser command from utillinux doesn't work. This is apparently a PAM issue. Original message on the mailing list: System: 16.09pre79463. (Flounder) Nix version: nix-env (Nix) 1.11.2 Nixpkgs version: 15.09.505.\nThe util-linux (2.27.1-6ubuntu3) package in ubuntu xenial ships with the following pam.d config:\nBoth files are shipped by and respectively in the original package source. It should be possible to integrate them in our own package. Fetch and unpack.\nThe following line in the NixOS config appears to solve the problem. But it would be nice to have it fixed in the package itself. Also I know next to nothing about pam so I might be doing something very wrong."}
{"_id":"q-en-nixpkgs-7401c77f1e90484f8b64b454dee9a7785488022d0dda90c8caaf927c999e7d62","text":"A clear and concise description of what the bug is. Steps to reproduce the behavior: -p nixops ... ... it builds! error: builder for '/nix/store/48sdc9gb26pb3prykylvy644mmsma0km-python2.7-ruamel-yaml-clib-' failed with exit code 1; <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Maintainer information:\nBuilds on linux thanks to .\nstill fails:\nI think the issue is tracked in I do not understand what the state of nixops is. is an issue about NixOps 2.0 with all checkboxes ticked. Further down in the thread there is a link to a board called . So what's going on? Are the things in that board release blockers? Which version of nixops should normal users use today? The project does not seem to make it clear.\nAnd even with that insecure package enabled, the fails with an import error.\nNixOps 1 is effectively end-of-life. That issue does not include critical usability problems present in master. To quote my response to another issue: All maintenance to the NixOps core currently happens on NixOps 2 (master / pre-release), because of the project's limited resources. NixOps 1 relies on python 2.7, so maintaining it is not feasible with the NixOps project's limited resources. Please consider helping with maintenance and testing, or donate to This fund currently supports reviews and fixes on master. Donations greatly increase our ability to get things merged and work towards a functional NixOps 2 release.\nIs there a current way to get oneself an overview of what's working in NixOps 2 master, and what isn't? Is it I would like to contribute testing it, but need to know: what the migration path from nixops 1 is (can I continue to use my existing state files?) whether there are dangerous unaddressed issues (will it wipe my infrastructure?) if there are key unsupported parts that nixops 1 has (Hetzner support, EC2 support, Route53 support)\nYes, this is more up to date than the previously linked issue. I'll have to refer to the issues and PRs. For the core, that's The documentation is still lacking in this regard, and some more work on the legacy state storage backend or other state storage code may need to be done, e.g. In my experience, NixOps 1 state files can be picked up by NixOps 2, but the opposite may not be true. Unlikely, but check the issues and prs of the plugins you're using. The plugins that are responsible for these parts should continue to support whatever those parts used to support, and they are available. IIRC we have two plugins for Hetzner now, one using a newer API. I have not worked on any hetzner related code.\ndon't you dare.\nThe build passes, if you permit that insecurity, now.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-a2c7c4a6a15216aab23b467f9e19ff347cf1c68b96dc8e590203290e51310f40","text":"Launch GNOME Music 3.18 There's no artwork at all Log: System: (Flounder) Nix version: nix-env (Nix) 1.11.2"}
{"_id":"q-en-nixpkgs-0d46ce628edc7edffc4f50bc16e42bdb9f2495d2e9e438244bef9b8b20b852fc","text":"When is enabled, my printer is not recognized by . Initially, when the option was not enabled, my printer would immediately get recognized upon plugging in the USB, show up in the Printers section in , and printing also works fine. After enabling SANE, the printer is not recognized, but the in-built scanner gets recognized and works properly in apps like . After disabling it and running , the printer is again immediately recognized, but obviously the scanner doesn't work. Configuration: Both the printer and scanner must function and get recognized properly when the above-mentioned configuration is applied. SANE did not give any errors Error after enabling : After disabling it, the printer gets recognized: Based on commit history for the SANE and modules Printer: CANON Pixma MG 2570 Driver: CUPS+Gutenprint v5.3.4\nthat is extremely puzzling, as sane is enabled or disabled though an environment variable that cannot be changed by a mere nixos-rebuild switch. you need to reboot or to logout/login to effectively toggle it.\nYes. initially, in order to enable the scanning functionality, a reboot was required. But the connection to the printer was established by simply disabling the option and rebuilding\nWithout lookling deeper into this, are you sure it’s not just a random coincidence? Were you setting up printing and scanning in the same session? Maybe you changed something else in-between.\nJust tried it again today, same results\nI have the same issue.\nPretty sure it's a permission problem. With cups enabled: With sane, saned and cups enabled: After a reboot, permissions are back to the first case. I guess there's a problem with sane's udev rules. Update: Sane or saned define udev rules that change the group of printer device nodes. I've tried setting , but libusb seems to do extra permission checking beyond the device nodes, so this doesn't allow cups to print.\nNice find! Do I understand correctly that when the device is owned by lp scanning does not work and when it is owned by scanner printing does not work?\nHello! Yes. From my understanding of , generates udev rules for every \"scanner\" device that changes the group permissions of device nodes to \"scanner\". This does not work for multifunction devices, as Cups expects device nodes to be accessible via the \"lp\" group. I've created a quick fix for this by overriding the derivation to delete its udev rules. Note that SANE is set up for device nodes being owned by the group (seems to be implied by . I've worked around this by overriding sane backends to delete the udev rules: This is a temporary fix, but I don't know what the \"proper\" way to do this would be.\nOk so when the device is owned by lp, it also works with the scanner. Reading you it looks like an issue with upstream udev rules, I wonder if filing a bug upstream is warranted.\nSeems to be a known issue:\nLooks like is the proper solution. would you mind giving it a try? It's hard to test without affected hardware...\nTurns out the command was never run, and the rules were just copied from the sane source. requires a search path to find description files. After patching via rules are indeed generated, and the owner isn't changed.\nSounds nice. Would you mind opening a PR?\nI went ahead and created , due to lack of response.\nthis was fixed, but incorrectly, in . A fix to the fix is in , it would be nice if people affected by this can test it.\nStill works for me, tested printing & scanning with the following overlay:\nClosing since it sounds like some combination of and fixed this, but I'm happy to reopen if I'm misunderstanding."}
{"_id":"q-en-nixpkgs-51bd464a3e69069125b14aba2c272c0a81c0b23c3c7b19928d36640200d28576","text":"unbound-full cannot be built on the unstable channel error message: Steps to reproduce the behavior: build .#unbound-full It should obviously build cc Maintainer information:\nThis happened in the fixupPhase, maybe we should remove lines?\nI've put some information on"}
{"_id":"q-en-nixpkgs-d7fbffacb18c8af19d6ef1da19ac7a5b61fa1acf5675c622b0a67aa2fd863c41","text":"The option does not work on aarch64-linux due to See also It seems that the current issue is that does not build on aarch64-linux. Steps to reproduce the behavior: NixOS on aarch64-linux. the option. the failure. aarch64-linux support n/a See also <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Maintainer information:\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\nIt looks like is only pulled in when : Maybe the assertion could be relaxed a bit?\nBump. Please change this so that we can at least use the guest tools in headless mode as mentioned here:\nCan you submit a PR and change the assertion message? This makes sense.\nor could you give a try?\nAny suggestions on how to update the assertion message? Am not sure if it should also be conditional.\nCan you tell me how to single pick this PR in my configuration? Define an override?\nYes. There are multiple ways to do this. So finding the simplest or easiest way will depend on your current setup. How do you define/import nixpkgs in your configuration? That'd be one way to do it. AFAIK you'd need to add the vmware- file to (), and bind a new nixpkgs reference to my branch, where you'd import that module from.\nAFAICT this is also a blocker for creating a graphical ISO for :\nSeems to work on aarch64 for me:\nDoes not seem to work for me: ! ! Rebuild: ! Assertion is in ! (Sorry for screenshots, but copy/paste does not work yet ... as it requires the vmware guest tools ... )\nAny progress - I am a begginer in nixOS. So the way to learn is setup a VM But then I hit this. What are the exact steps to use this PR? This is not just a mouse issue as the previous comment not it stops cut and paste. It also shops the use of shared folders so a fix is needed for the headless case as well.\nI would love to have vmware guest on aarch64 too. I'm running VMWare Fusion on my M1 macbook. I tried the above, of disabling the vmware- and adding my own with the assert modified to accept Aarch64. However it not building x11 mouse is one issue, but even with headless = true it still fails; in this case vmw_vpscsi fails in the kernel build: !\nthat's a module that comes with the kernel and not vmware. It looks like it's not supported under non-x86 architectures: so we should remove it from the module if not on x86\ncan you test ? I don't have aarch64 but this might work.\nMagic. It built successfully and as soon as nixos-rebuild switch completed my X desktop resized to match my monitor (automatically). I just used my own local vmware- (disabling the system one), in which I made the two changes that are in kjeremy's PR ^ Thank you.\nCould you elaborate how you did that? The furthest I got still gives me"}
{"_id":"q-en-nixpkgs-5020c8f05012508cbd1a9ff20c1ef1f39ec3ddb5f29c3508f3296ba5cca59a93","text":"Get a mac with facetimehd support and the following the laptop lid the lid back (might also need to press a key) I'm running unstable and assuming its broken on 16.03 too. System: (NixOS: , Ubuntu/Fedora: , ...) 16.09pre79453. (Flounder) Nix version: (run ) nix-env (Nix) 1.11.2 Nixpkgs version: (run ) \"16.09pre79453.\"\nIt is broken since since the module is using which is obsolete and can't handle compressed modules.\nCan you report if fixes this?\nyes switching to kmod works (only tested facetimehd). Are there plans porting this over to 16.03?\nYeah, I assume we should port it &ndash; it doesn't seem likely to break anything and it does fix some bugs.\nOh hrm. I missed this issue conversation and put in this PR: I reckon we should backport , and\nThis is fixed in 16.03 and master. can you close?"}
{"_id":"q-en-nixpkgs-ed79898ddf4e1e09f17109e404043a131047f5f0cac01dc35ddd094e773045a4","text":"libcanberra is not included in of thunderbird and firefox. This prevents both applications from playing gtk sounds (for instance when receiving an email on thunderbird). Steps to reproduce the behavior: any instance of firefox And in the output of strace you will see that firefox doesn't try to open the lib dir of libcanberra in the nix store. For instance on my system: libcanberra is supposed to be in the of firefox and thunderbird wrappers. AFAIU firefox and thunderbird open this library at runtime with dlopen and that's why it's not in the generated . <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Maintainer information:\nYou are correct. It probably should be to the wrapper. It seems is pulled in by . Not sure what it is doing or if it is needed but adding to should address . On another note, there ought to be a tool that searches Firefox's source code for all runtime shenanigans. Also the wrapper should probably be expanded so that overriding packages into is possible.\nThis bug still exists. Is there any idea?"}
{"_id":"q-en-nixpkgs-f798c4969484eb32d8c7de4dde01f881e50a9c48b43bac914d31b2afc5428f31","text":"Running fails with the following error: The issue is here: Before the 2.8.0 update, this line used to say So, flutter copied the cache to , which was fine. Now, however, it tries to copy the cache to , which doesn't exist, so flutter throws an error. Steps to reproduce the behavior: do any flutter project `flutter run Flutter builds the app and runs it on the device Please run and paste the result. I cannot run it because I use flakes, so it just says .\nMy bad, it should be fixed by\nthanks!"}
{"_id":"q-en-nixpkgs-a3c2232c6f9456ff4750772457c63e12cab2cc03cf9ec86a8b132fb1c90371b0","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on github to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: roon-server <!-- The current version can be found easily with the same process than above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 1.8-850 desired version: 1.8-880 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-9d7c88cad4fb177d2e252bf2eb825bd1ce86566f2e7f5e8604d21176758624b8","text":"mediathekview might be affected, the latest version 13.8.1 patched the distributed log4j version As it also updated compatibility to Java 17, and no working Java 17 distribution is in 21.11 or master atm, I'm still working on how to best address this.\nSee PR Full update depends on updated Java 17 , manual patching might be needed in the mean time. As it is distributed as a binary and log4j is packed into the main jar, I'm a bit stumped at the moment how to approach this.\nMight work\nSee PR for a fix"}
{"_id":"q-en-nixpkgs-2aeabe3c40dade80b9955aff04b52f437da5dcc8c3e2f21bc1b5094bd9a5bd9d","text":"Upgrading fails, seemingly because the existing systemd version is newer in 21.11 than unstable. Steps to reproduce the behavior: NixOS 21.11 21.11 channel, add unstable channel NixOS upgrades from 21.11 to unstable. is the commit upgrading systemd to 249.7, cherry picks it to 21.11. Adding the flag to the invocation would also solve this issue, albeit leaving the newer EFI files in place. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\nI think the corresponding PR is\nAs pointed out above by this will be fixed when staging-next gets merged ().\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-bc23109d20947296c93cb64009f5fd980d985d7375376afbc3635c3dc003832e","text":"However, the following input is is rejected as \"*: incorrect address specification\": I expected kresd to listen on fe80::1 on port 53 on the ofborg interface, as the documentation suggests should be possible. I don't mind that it is impossible, but the documentation for this option isn't correct anymore. nixos-unstable-small, rev\nWell, it no longer uses systemd sockets anymore, and instead there are a few in-nix regexps that convert them. It might not be perfect. Either way, the interface spec should surely be part of the IPv6 already, i.e. &ndash; and I think that should work (the generated lua config looks OK).\nAh, I see that systemd is weird here, at least in docs. So I don't know what we'll do with it. Either shuffle it in the in-nix regexps or document as a deviation, I suppose."}
{"_id":"q-en-nixpkgs-e34ad2bd63af9377141470039171f154603eaf7d78ee4eee03868302bd2c19c1","text":"I get the following collision: What I don't understand is why this collision is happening in the first place? I'm installing this package into an empty profile!\nThat's a multiple-output problem. The derivation creates two of them, both containing the file. When installing into user profiles, currently the semantics is that all outputs are installed, so this creates the collision.\nVladimír Čunát writes: What's the reason for these semantics? I would expect only 'out' to get installed by default.\nWell, as for build inputs, we normally want all to be available, because you need include, lib and possibly others. should know details. It hasn't been used much yet, so I suppose we still have to figure out how to do these best (and have the most usual setup as the default).\nWhat I haven't figured out in these cases: why installing attempts to install both and .\nI am getting this same error message. Fresh install of Nix 16. Could it have regressed? This happens when I run nix-env -iA nixos.openjdk but not when I dun nix-env -iA Though those both say they point to openjdk-8"}
{"_id":"q-en-nixpkgs-ac3a731c5022ce90659a1b7946969ea8aca294513c3e6dffcb9c90409658a3b0","text":"I think it should be marked as broken to avoid frustrating other people. This package is impossible to build without more than 4G address space due to linker memory exhaustion. It can be built only on 64-bit hosts (the same as Firefox). And that permanently breaks Gnome 3 on i686. It was possible to build it in the past using local modifications, though.\nwhat modifications did you used in the past?\nOnly very few to some Gnome dependencies just to build Gnome itself, like . AFAIK, there were no modifications at all to webkitgtk. It just required huge swap and to free as much RAM as possible, e.g. running nix-build with only one thread. Last December I've tried to install NixOS 21.11 and it was impossible to build because of a linker exhausting all memory address space. I've tried both ld and gold. My investigations into other Linux distros revealed that they used cross-compiling tricks to build 32-bit webkitgtk on 64-bit host. I've spent a week before giving up. Right now I'm using only x86_64 arch because of that issue. I suppose it's impossible to fix on native 32-bit host just like Firefox.\nI tried different configurations and was able to complete the building without any memory issue by setting separateDebugInfo to false.\nI will now try to build firefox. Edit: it is markerd as broken on 32 bits.\nInteresting. Thanks for trying this. It was very painful to experiment with different configurations as it required too many hours to build on my computer. AFAIK, Firefox had a different issue which didn't involve linker. It was probably a compiler. Do you think it's sustainable? I mean, will it still be possible to build next year? Because a year ago it didn't require that trick either. I'm quite pessimistic about browsers on 32-bit architecture as is already evident from Firefox :( Also, there were similar troubles with Haskell (Pandoc) and Qt. Considering that Hydra doesn't have most 32-bit software in a binary cache and that it's often required to set in order to build it, it's really a huge pain to install NixOS on i686 arch.\nworks great for me. Should we add a conditional for this derivation?"}
{"_id":"q-en-nixpkgs-41c3bf3e1df6632726d12a9af093b782f927b7790ebfe52308b1bda89b9965f2","text":"When attempting to add a Vim plugin, fails. It updates without error. I believe previous developers have this feature enabled in their system/user Nix configuration."}
{"_id":"q-en-nixpkgs-79cee2db9c0817dccdcd9e96ae2c26e3ab28dc3452de3cef55ce8390a1492c2b","text":"A follow up to It seems like is not getting updated when tmux sessions begin/end. Relevant links: Steps to reproduce the behavior: and note the results. a different shell, open a new tmux session: exiting tmux, run again and note that the results are the same. Tmux user sessions to show up in , , and , as is custom in other OSes. n/a n/a <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result."}
{"_id":"q-en-nixpkgs-36d69ee68d3c0891bd75f11a22034d010324e1b08dd1d97a1393bcb26dca6e11","text":"I'm trying to start the service fusionInventory but it fails with the same error as , running the system command in a terminal also trigger the issue. Steps to reproduce the behavior: part in , this is the minimum to have fusioNInventory to start, the server doesn't need to be real, it fails before trying that. I have a real setup that also fail with the same error, I tried this one to be sure it's reproducible by everyone. note that this command line without works: the service is running fine\nSolved in after debug with upstream"}
{"_id":"q-en-nixpkgs-1cc060c9feab529371bb7477a8e20dc0b2caaa6821a8d6132ebeb206d25e21e4","text":"After updating to the latest NixOS unstable channel, when running , the process will fail when building fcitx-libpinyin. Can repro this in a fresh installed NixOS using , adding the following lines: channel: : process will fail when building .....-fcitx-libpinyin- with exit code 2. The rebuild process should succeed without failure. !\nUpdating to 0.5.4 seems to fix it:"}
{"_id":"q-en-nixpkgs-640d9ae1365e0df306ee0ebb54337191b252e0bc185c2afc9c785d6588b2e05a","text":"Blender 2.93.5 Hallo When I start Blender now and set under “Render Properties” → “Cycles” Blender crashes completely. Also when I go to “Edit” → “Preferences” → “System” Blender crashes immediately with this Error: What does that tell me now? My GPU configuration in the clinfo: nix run -c nix-info -m:\nI think this might be the same issue as described in"}
{"_id":"q-en-nixpkgs-30c3fb1e0e498a2c7b17f241c1adad7ab427c9f8ac4caad491db6a3165550155","text":"I found a disappointing issue with NixOS on AWS EC2 - absence of support for Enhanced Networking (SRI/OV) out of box. See docs at Here are separate issues: is outdated on 16.03 (less then recommended 2.14). Amazon Linux AMI is better then NixOS here amis are built without --sriov flag. Because of 1) this makes starting spot instances with --sriov trickier As I'm not an expert in AMI creating, instance management or kernel module packaging, I won't do pull requests on this topic. But I'd like to share some success with others. Here is the ixgbevf-3.1.2 kernel module package. it And here is extract from config to enable this module: My instance started and ssh works. Now I have this: Ping\nThanks, I wasn't aware of Enhanced Networking. I'll look into it.\nAdded the driver to the Amazon config: , . Will need to rebuild the AMIs though (and make NixOps enable Enhanced Networking automatically), and add the driver for instances."}
{"_id":"q-en-nixpkgs-24f39b0d9498ef72ecb82cc74825fce7c2214e8eeb6167519d064017047aaf8d","text":"While going through the documentation I came about a that explicitly requires matplotlib with the TkAgg backend. I tried to generate a nix-shell with matplotlib and tcl/tk and tkinter. However, the script does not execute and instead I get the following error message: In a simplified form: Start nix-shell:Start Python:Try to import matplotlib, configure the backend, and import parts of Tk:Observe the following error message:System: NixPkgs on openSuse 13.2 (Harlequin) (x86_64) Nix version: nix-env (Nix) 1.10 Nixpkgs version: \"16.09pre83147.\" Matplotlib Maintainer:\nJudging from the matplotlib expression we don't build against tcl/tk.\nI made an attempt to build matplotlib against tcl/tk with the following patch: Unfortunately, compilation fails with the following error message: The linker cannot find the tcl/tk libraries. The linker flags are quite off: They include and the like. This must be due to matplotlib's own build script, as no such flags are set in my bash environment and is executed with the flag anyway."}
{"_id":"q-en-nixpkgs-8c2bfcd96df50b937a8ec22bd48d1ab6d4c560c3b7d8633b398b9b00e109ace6","text":"I would like to use the following option in in : When running the command I get following error: This is because are appended after the option. It would be better if we switch the order of these. Steps to reproduce the behavior: following configuration to execute the command will fail with following error: The program should be exectued with accepting all"}
{"_id":"q-en-nixpkgs-ba2ddf113c5e384af8689c6fe8f666cbad2fdfed6ab5236eea3790cdaacebb43","text":", CVSSv3=9.1 (nixos-21.11, nixos-unstable) CVSSv3=9.1 (nixos-21.11, nixos-unstable) CVSSv3=7.5 (nixos-21.11, nixos-unstable) An issue was discovered in the DNS proxy in Connman through 1.40. The TCP server reply implementation lacks a check for the presence of sufficient Header Data, leading to an out-of-bounds read. An issue was discovered in the DNS proxy in Connman through 1.40. forwarddnsreply mishandles a strnlen call, leading to an out-of-bounds read. An issue was discovered in the DNS proxy in Connman through 1.40. The TCP server reply implementation has an infinite loop if no data is received. Scanned versions: nixos-21.11: ; nixos-unstable: . Cc\nUnstable - 11 -"}
{"_id":"q-en-nixpkgs-0d553afd8e34555e412b456b1cd25f993a586988a5bb10eb72ea464d6d492901","text":", CVSSv3=7.8 (nixos-21.11, nixos-unstable) CVSSv3=6.5 (nixos-21.11, nixos-unstable) CVSSv3=5.5 (nixos-21.11, nixos-unstable) CVSSv3=5.5 (nixos-21.11, nixos-unstable) CVSSv3=5.5 (nixos-21.11, nixos-unstable) There is an Assertion `scalinglistpredmatrixiddelta==1' failed at in libde265 v1.0.8 when decoding file, which allows attackers to cause a Denial of Service (DoS) by running the application with a crafted file or possibly have unspecified other impact. An Incorrect Access Control vulnerability exists in libde265 v1.0.8 due to a SEGV in An issue was discovered in libde265 v1.0.8.There is a Heap-use-after-free in intrapred.h when decoding file using . A stack-buffer-overflow exists in libde265 v1.0.8 via fallback- in function putepelhvfallback when running program . An issue has been found in libde265 v1.0.8 due to incorrect access control. A SEGV caused by a READ memory access in function derive_boundaryStrength of has occurred. The vulnerability causes a segmentation fault and application crash, which leads to remote denial of service. Scanned versions: nixos-21.11: ; nixos-unstable: . Cc\nSee"}
{"_id":"q-en-nixpkgs-f52ad7abe32b6fa5f59c6c02072a5612d3976e9100cafaa862ca7fce66dbb204","text":", CVSSv3=6.5 (nixos-21.11, nixos-unstable) A NULL pointer dereference issue was found in the block mirror layer of QEMU in versions prior to 6.2.0. The pointer is dereferenced in mirrorwaiton_conflicts() without ensuring that it's not NULL. A malicious unprivileged user within the guest could use this flaw to crash the QEMU process on the host when writing data reaches the threshold of mirroring node. Scanned versions: nixos-21.11: ; nixos-unstable: . Cc Cc\nThis is fixed on master since we have QEMU 6.2.0. It's not included in 6.1.1 though, so needs to be fixed on stable. Probably after since it'd be a merge conflict."}
{"_id":"q-en-nixpkgs-f65b35d8f06a45b5e9f18e68ff99af25a28c00ba0da7b18048215a23e7440dcf","text":"I want to use (providing ) for it's Debug Adapter support. Particularly it seems like Helix can utilize it with some early debugging support. However, it doesn't seem to work. After adding that to my nix shell, and entering it, and trying to run it, I get:\ncc: since I'm running your latest PR bumping this.\nI suspect that vscode-lldb's fork of lldb needs to be updated?\nOh yeah, that would be because the adapter is patched-inline in the extension process and the unwrapped, useless, adapter is passed thru instead.\nI have some work that fixes this: just uses regular lldb now, we can drop the \"fork\" with the \"rust-enabled\" check tries to exec-itself in certain scenarios and must be patched so that it properly calls its wrapped version But most of all, it seems like is ... just built as part of already. I'm not even sure if there's a huge point in bothering with and in general. I really wish this were less messy. was a hassle to figure out, has a bad name, the repo that you find online is old and says darwin-only. yeesh.\ni got hit by this, but now it says\nActually only resolves the issue. The original issue of missing still exists somewhat. Yet a working binary can be found after that PR at Help is wanted to resolve the path issue of the adapter sub package :smilingfacewith_tear:\nwhy was this closed? I don't think it's fixed."}
{"_id":"q-en-nixpkgs-8e7624cd66432a4913e872891df8636f2097974e8e231e68815cab1bde281cfd","text":"pivy fails to build due to being unable to find some files Steps to reproduce the behavior: try to build pivy from nixpkgs HEAD"}
{"_id":"q-en-nixpkgs-849e9056777800f653cf56f1e2756be73b0fb29e9e04ef6cda906dd505767096","text":"I am encountering the following error after launching Pinta 2.0.2. I thought initially that it was a problem in the Adwaita theme itself, but whatever the GTK theme I use, I always have the problem. I tried on two different machines and had the same problem with both. NixOS with unstable channel Ubuntu + Nix 2.5.1 with unstable channel I did not have this problem in the last packaged version of Pinta. Maybe it is there since the 2.0.2 upgrade? Pinta should launch without this error. ! Please run and paste the result.\nThis looks like exactly the same problem discussed on Except that it should be fixed. Not sure what is happening here, and I don't have access to a Linux machine right now to debug. Maybe can help?\nI also have this issue. It also causes the toolbar on the left to not appear.\nCan someone show me the contents of the directory inside the package? It should have a directory/file somewhere, inside the icons directory (I don't remember the exact path). If this is missing, this is probably the issue.\nIt is missing:\nHuh... I think I can blindly patch this issue, however someone will need to test on the PR.\nWait, I was wrong, it is there, but not in :\nIt shouldn't be in , it should be in . This is the issue. This can be probably fixed later correctly by setting some setting on , but since I don't have access to a Linux system for now I will just duplicate the assets for now. Later on I can do a proper fix.\nCan someone test this PR please:\nWaouh that was fast. Thanks a lot, both of you for your reactivity!"}
{"_id":"q-en-nixpkgs-9115912f8f36d47d00d498227a572f62cf0395566b55fca6e52cf55c39eb5380","text":"From Currently nixpkgs has as an . We should either have as a nix list of enums, or just a string to allow the user to specify a zsh array."}
{"_id":"q-en-nixpkgs-e44d30508cc134b7576f40a7c37d4d797aa69df98f02e967f0a3591d16cde018","text":"Project description Cider is an open-source and community-oriented Apple Music client. Metadata homepage URL: source URL: license: GNU Affero General Public License v3.0 platforms: linux, darwin\nThey ship an AppImage which I use in my own config packaged as follows:"}
{"_id":"q-en-nixpkgs-efc2a007a300820e3e07a35a2cf665b0cf4628e5a0b6f2df5fd88dc900aaa603","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: element-web <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 1.10.4 desired version: 1.10.6 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Myself i am waiting for an , ... please keep an eye on the release."}
{"_id":"q-en-nixpkgs-486e3920f9d1ab2186439cac7edc25b1091e16e58e06e7b558349de302b5cf92","text":"I just found out about wrapGAppsHook by examining a gnome package, but there really should be documentation about this in the Nixpkgs User Guide (a section about Gnome would be nice, which would show the right way to package GTK3 / GTK2 glib based apps). Also, perhaps there are other things I missed that one should know when packaging Gnome Applications or GTK3/2 apps? /cc Gnome maintainers\nI'd like to know as well. Currently receiving this error: On one of the packages I'm trying to put into nixpkgs. Apparently sometimes you need for buildInputs. So apparently all you need to do to use this is:\nThe error occurs when you do not have any directory containing in variable. One of the things does is (which usually contains the aforementioned file) of dependencies to . (Actually this happens in cooperation with via variable.) Icon themes work in the same vein, only is used and it is populated by which is propagated to (). (XDG fallback theme) should be probably to any GTK application that uses icons (I think that is every GTK applicaton nowadays due to CSD) but some applications require Adwaita because they use icons not in hicolor. Not sure if this functions correctly though, at least it should not according to my mental model of how icon theme selection works; I will need to read the relevant XDG spec before I can finish the documentation.\nJust thought I'd leave this comment here for future reference: if you get an error like for python gtk applications, try adding to propogatedBuildInputs!\nIt should go to , though. But be wary of"}
{"_id":"q-en-nixpkgs-d529746ee3cae91cf7d91ce4dbe4e6aaaf40857193a6541074784e3f8d900e78","text":"Project description Layan kde is a flat Design theme for KDE Plasma desktop. In this repository you'll find: Aurorae Themes Kvantum Themes Plasma Color Schemes Plasma Desktop Themes Plasma Look-and-Feel Themes Metadata homepage URL: source URL: license: gpl3 platforms: linux"}
{"_id":"q-en-nixpkgs-defc4c96aefb9c61712bc195209da312ccc479b73d038acd31dfa40c283f6441","text":"When attempting to install , installation fails with the following error: As noted in , the call to is incorrectly using a capital B in the flag. It should be . The error coming from the matches the error when running the same command by itself: I searched for an existing issue which reports this, but didn't find any (aside from the comment linked earlier). Slack is installed successfully. N/A N/A <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\nCan you make a pr fixing it? I can merge.\nI made a commit to do that in my fork here: However it still fails during the , this time with . This was also reported in but only on . However it seems the same issue is also present on and I'm not sure where to go from here on fixing it. I can run the same command from my terminal successfully. Is there anything different about running during the that might cause this?\nI think I figured out the problem. writes to , but this isn't accessible to the user running the script (). I'm able to reproduce the problem now if I try to run the command as that user. Is there some way to make this file writable by the user as part of the ? I'll try to do some research on how other packages set property list attributes on darwin to see if there's some other workaround."}
{"_id":"q-en-nixpkgs-5db51745a3326a4b7520d3129105044482e2046dd0a77540128805ef10bbd9c1","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: blender <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 2.93.5 desired version: 3.1 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-6433eb8bba19eaea0c0afa989d0ed7d2f3706dc8623ff358d4eee640d647239b","text":"Qtile is missing all the ffi modules so one can't use the pulse widget Just install qtile and add a To the config There is a script that has to be run as part of the packaging process takes care of creating the ffi modules, not sure if that will be enough or patching to change the paths will also be required.\nProblem still here, and only for this module"}
{"_id":"q-en-nixpkgs-32ebd7c35c595a95dcd1dc76ae94695b3ae52e179791db712de650fdaf3a4374","text":"Obsidian is unable to run (non-nixpkgs) git commands, due to being prefixed with . Steps to reproduce the behavior: non-nixpkgs from nixpkgs in obsidian the command Git syncing properly <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nWhy don't you use git from nixpkgs (e.g. )?\nThat does work, although I'd prefer not to install git with on manjaro. I have other workaround that i use, i simply wanted to report how seem to break subcommands.\nJust updated, and this also affects electron 16\nWe could consider prepending to in the derivation (the start script in the output).\ncan you guesstimate how many percent of obsidian users have obsidian-git plugin installed?\nOf all obsidian users: few. Of obsidian&cap;nixpkgs users: maybe 20%?\nPossibly related:\ncan you please check whether the fix in fixes this issue for you?\nSeems to work!"}
{"_id":"q-en-nixpkgs-2a4bc963dbe3c1b68d920ef906522eb4dfe9382ada2dd41072f4b39c1ad048e7","text":"Password hashes, while less sensitive than unhashed passwords, are secrets. We should therefore have a option in that avoids both placing the hashes in the store world-readable and storing the password unhashed.\nThis is a common misunderstanding due to the bad naming of these options: → string of plaintext password → string of password hash → file containing password hash So, what you want already exists and it's .\nOh wow, indeed. Thanks! Given that the option is named so confusingly and that its description starts with \"The full path to a file that contains the user's password.\" I think we should keep the issue, pending an option rename and a documentation improvement."}
{"_id":"q-en-nixpkgs-1f92666742a57733c6f9b01c5f6cb7510128de7c12ee0d7cf29ca1f1b7e83d0a","text":"yaru-theme fails to build on the latest nixos-unstable Steps to reproduce the behavior: fails Buiild succeeds ! `"}
{"_id":"q-en-nixpkgs-02a961e7d214689e3ed57b2a06b3f911eafb8e9f783e3f0b7460a20a237d61ae","text":"QEMU VMs using the profile that have a in their hostname were broken by A clear and concise description of what you expected to happen. If applicable, add screenshots to help explain your problem. Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result."}
{"_id":"q-en-nixpkgs-636986096f9efc6226e8e67105adc4dbc2bee7c171b8830272b816b9fffe2ab2","text":", CVSSv3=7.5 (nixos-unstable) Vyper is a Pythonic Smart Contract Language for the Ethereum Virtual Machine. In version 0.3.1 and prior, bytestrings can have dirty bytes in them, resulting in the word-for-word comparisons giving incorrect results. Even without dirty nonzero bytes, two bytestrings can compare to equal if one ends with because there is no comparison of the length. A patch is available and expected to be part of the 0.3.2 release. There are currently no known workarounds. Scanned versions: nixos-unstable: . Cc"}
{"_id":"q-en-nixpkgs-5fd8f346f51f0c9abbf14ed0398e1a998db1723a2f169ff374439d0a29c55ee9","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: coreutils <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 9.0 desired version: 9.1 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR.\nI'm already on it :) might take some while before I do the pr because I want to build stdenv first"}
{"_id":"q-en-nixpkgs-d2c5f81bfe782b11b63079217582430e4f5964c93ac4e56052fd3a820c341695","text":"Can anyone run successfully? When I try it, the tool almost immediately crashes:\nOn 6 February 2014 19:49, Peter Simons wrote: Same here. Seems to be happening elsewhere too:\nSaving to works around the issue for now.\nAny way this can be wrapped to happen automatically?"}
{"_id":"q-en-nixpkgs-8473122c4fc53deb5239a3a9a884f12363b739a6ff070d781ae35f5d85611678","text":"Build of failed on x86_64-linux as of . This is currently breaking maintainers cc: maintainers cc: Other cc: n/a nixpkgs at commit This issue was automatically generated by . that created this issue. Internal tag:\ncan confirm for and\nI'm trying to bisect this atm, but it's a truly massive undertaking since there's hundreds of commits between current master and the last successful build... plus there's lots of other irrelevant failures :/\nI was unable to find the exact commit that caused these failures, since it was interlaced a bunch of other commits that broke dependencies. Created upstream. We'll see what they say. We're not doing any patching of required dependency versions, so the test suite ought to be passing.\nI also tried bisecting. Due to the massive rebuilds, I canceled after 8 hours and did a bisect with . The bad commit is from , which is a merge from staging with changes to python and others: <details</detailsLooking at the there is a requirement for werkzeug < 2.1.0. When applied, the test mostly pass. The only failing test is: There is also a requirement for pytest < 7. Since this packaged builds with the latest nixos/unstable, which includes pytest 7.0.1 I am now testing with that version.\nThe updates usually breaks the package. I guess from 1.5.7 to 1.5.8. Edit: Was part of\nFurther investigation: When I apply the following two changes: downgrade to downgrade to sentry-sdk builds without issues. How should we go about this? Just disabling the tests could lead to malfunction of the sentry-sdk package (because they did correctly detected a regression in the SDK). Downgrading both packages Adding an override in for those two packages could lead to potential breakages due to dual imports of two different or imports in packages, which import sentry-sdk and one of the two packages. Obviously it would be awesome if upstream update their end of the SDK of werkzeug. This could take some time and sentry-sdk is needed by other packages.\nI notified the authors of the breaking commits and\nAnother breakage is pgadmin (tracked in )\nModule can't have overrides. Thus, till the issues are fixed by upstream the failing tests should be disabled.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-671a1439ca82272a88077468a41a0c85104326fd92909c19955b8c1313ed3770","text":"On current nixpkgs unstable, GraalVM 11 won't build (also breaking clojure-lsp, clj-kondo, babashka, etc). Steps to reproduce the behavior: macOS: It should build successfully graalvm17-ce fails for the same reason. GraalVM package maintainers: Please run and paste the result.\nMy bad, this was probably one change that I made on the previous refactor. This PR should fix it:"}
{"_id":"q-en-nixpkgs-80fbf86b557b254c7ec6a37faf5290a905433b9d2a3ed3353c036ad97d0e02e1","text":"Project description Add Khmeros TrueType font for Khmer language. Also, see All issues have attached PRs. Metadata homepage URL: source URL: mirror://debian/pool/main/f/fonts-khmeros/fonts- license: gpl2+ platforms: all"}
{"_id":"q-en-nixpkgs-98d36f4ec58ff4aa85af1f5bb658d89713c3dd705697a0e7e7a2de9b417fb9e3","text":"Project description Add Korean Nanum fonts: Nanum Gothic Nanum Barun Gothic Nanum Barunpen Nanum Brush Nanum Myengjo Nanum Pen Nanum Square Nanum Square Round Metadata homepage URL: source URL: license: ofl platforms: all"}
{"_id":"q-en-nixpkgs-984da43f164f11f04407bde6b594879181854afe50c689f83dac3907318f97b7","text":"Commit from broke haskell.compiler.ghc6102Binary haskell.compiler.ghc704Binary and therefore the packages haskell.compiler.ghc6104 haskell.compiler.ghc704 haskell.compiler.ghc722 haskell.compiler.ghc742 haskell.compiler.ghc763 which depend on them.\nPlease try . I test-built the binary GHCs but no further.\nAha. I relied on the update being API-compatible, but the first bootstraping steps don't recompile...\nI'm test-building the whole list above.\nAll succeeded on x86_64-linux."}
{"_id":"q-en-nixpkgs-059c066c193bb238629dec1fded29492a45bbffb678a10b1bf5332831307d747","text":"Krita fails to build with the following error Steps to reproduce the behavior: Build Krita should build. nixpkgs @\nIn libraw was updated. Switch to the old stable libraw020 to fix this build.\nInstalling kirta is showing like it's not in the cache, any idea why?\nYou're attempting to build the older derivation. And building it would fail. Newer binaries are available:\nIf you rely on channels, you might want to watch\nI'm using the \"\" and just updated it, and I get this: So, that means the PR is not in that channel yet? If it's that, good to know, didn't know that tool.\nYes, channels that are not -small usually lag by roughly one day."}
{"_id":"q-en-nixpkgs-20893d406585a002426fc5d51e22a14b6ce7af9990d709c7706806fb17c7b4bc","text":"The package is failing to build Steps to reproduce the behavior: The package to build successfully without errors\nI'm sorry, I haven't used in years - but also nothing with that package has changed in some while. I'd assume an upstream problem?\nI don't use but took a look out of curiosity. It was an upstream problem that is fixed by a few commits after the latest 2.4.0 release. There hasn't been a new release since then. Is the best thing to do here pull down that commit as an isolated patch or to bump the revision that's pulled down to that (untagged) revision?\nCan we use in the meantime?"}
{"_id":"q-en-nixpkgs-570b4d414bf48914dee4c371ccb2cef70affca91e923fda532db7824cffce11d","text":"scipy doesn't build on Darwin right now (trunk). Hydra: The error in Hydra seems to be some builder failure and not actually related to the issue. It however fails as well when I try to build on my Mac. We should get rid of both issues. Steps to reproduce the behavior: to build It builds I don't need this Python module myself but a lot of other packages depend on it. This is currently the package with the most dependants (322) that fails so it'd be nice to have this cleaned up before ZHF starts. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nThe too many open files sounds like it could be related to the very low default limit on open files on macOS. For me it defaults to 256, I think more recent versions have increased the limit but it's still lower than for Linux and the tests were probably written with that in mind.\nscipy builds for me on both x86_64-darwin and aarch64-darwin. $ nix-info -m system: host os: multi-user?: sandbox: version: channels(reckenrode): channels(root): nixpkgs:\nwhat does report for you?\nOk, then my hypothesis is debunked.\nCould this be due to the file descriptor leak bug in nix ? This would explain why it fails to build for some users but not for others. I believe the bug was fixed in nix 2.8.\nscipy seems to be build for me on , can anyone else confirm? Looks like it's still failing on .\ncan we just disable the tests on for now? as a note (or whatever) can bump the open file limit but you might then run into memory limits like I've experienced in\nI'm going to disable tests on so we can work on other issues for ZHF. If anyone wants to debug specific failures in scipy please reply to this issue. It currently causes hundreds of other packages to transitively fail.\nWonder if reducing parallelism would help."}
{"_id":"q-en-nixpkgs-fc6a8fbe9b55027e5c847420fd59917349ed6b418482aa7340626a339ce4ff23","text":"The issue is what the title says. By half, I mean and for my previously used from emacs-overlay to your system the eln cache in your home, e.g. to open emacs for the manual for by and inputing the eln cache in your home: , you will find together other eln cache. I think this is a bug since is builtin for emacs and it's eln cache should be distributed together with . Uses of will not need to native-compile builtin libraries themselves. does distributed together with from emacs-overlay I used before. It's . might be related to the bump of 28 , so\nThe difference between the two versions is that one is built from the tarball, while the other is built from Git. The tarball includes byte-compiled () files but not native-compiled () files. The build notices the files and does not rebuild them, but native-compilation is a side-effect of byte-compilation and so the files aren't built either. (Almost all of the files you see for the new version are trampolines, there are only 3 non-trampoline files.)\nideally we'd run , which is the expected way of deleting such things (or perhaps even ). But the tarball only has a , so we'd need to run , then , then again, which is super ugly. Hopefully deleting the files the naïve way is enough?\nI think maybe we should just skip the tarball entirely and fetch it from git instead.\nThere are other side-effects caused by skipping the tarball. This claims, for example, that bootstrapping is only expected to happen to emacs developers. Part of bootstrapping is converting the Japanese input method dictionaries, and this takes some amount of time. They're in the tarball at , but are absent in the git checkout.\nThat's a good point, I would still rather inflate build times a bit and exercise bootstrapping as a part of our regular build though to completely avoid these types of issues.\nI'm hesitant about building things in a way that upstream doesn't expect, unless necessary to adapt for nixpkgs. For a related example, using the repo instead of the tar distribution for Ruby caused a subtle problem where the commit hash was not propagated, which broke the bootsnap gem's cache invalidation (fixed in , by using the tarball). The real question here, I think, is how upstream expects native compilation to be used if their tarball doesn't support it."}
{"_id":"q-en-nixpkgs-a5d33a9a14bf8b412ca298a94b621ea663457250480c590c7b74aa0d1da71377","text":"On startup, Tor Browser doesn't start if some variables about Wayland aren't set. Steps to reproduce the behavior: and (if set) Tor Browser in a terminal and see the error message. Tor Browser should start. This issue was caused by . currently does not fix the problem. In the meantime, an easy fix would be : <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Notifying peoples from the aforementioned PRs instead. Please run and paste the result.\nI can confirm this issue\nthanks !"}
{"_id":"q-en-nixpkgs-a4fe6c22249cfb542793af24ce6d4e3bbf5832eaba41784380247365ce54cb01","text":"The option, which enables by default on high-resolution displays like my 4K laptop screen, is supposed to set the console font within the initrd before displaying the LUKS password prompt. However, with , this does not happen until stage 2, so the prompt is tiny. Enable , , and . The console font should be switched before the LUKS password prompt, not after.\nRelevant code is probably that it sets . should then properly set everything up if it can find its files in :\nDoes it work when you do ?\nNo, it can’t, because points to a font file that’s not present in the initrd.\nThis works though. It’s a little jarring in that the font is changed about a second before the framebuffer resolution is changed, so it’s temporarily huge. Ideally those would happen simultaneously. But at least they both happen in time for the LUKS prompt."}
{"_id":"q-en-nixpkgs-646f8aed06559982dd82020326283b93a8552bb977d83c4d81bbe535cdb226fc","text":", an application that simulates user input to keep the screen awake when pressing gamepad buttons, relies on on Plasma Wayland to do this. However, isn't installed (or at least, it's not in ) with a default Plasma installation, making not work: (Among other errors) Steps to reproduce the behavior: Plasma (), without \"installing\" ; a Plasma Wayland session; ; 'll see the error message, and the program won't work. There is no error message and the program works (which is what happens by installing ). I don't know if the best solution is to add to Plasma or to . I'd vote for the former, since (if I'm not wrong), is already installed with Plasma, it's just not in PATH, and I don't see the harm in being there. Still, that's why I opened an issue. By the way, can't be installed \"alone\". NixOS says: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result."}
{"_id":"q-en-nixpkgs-3af007a9c2bdd32d464c2b9dc720c18168d4e8a10d6570e2b496ff3b6c5f6ff7","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: qtcreator <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 5.0.3 desired version: 7.0.1 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR.\nVersion 8.0.0 is now available."}
{"_id":"q-en-nixpkgs-8b99ec40de99cfa162d9277f405bd95037a4d89604b281a317aacf1db3ef6c24","text":"I notice zoom-us from nixpkgs-unstable gets killed by SIGILL shortly after the application window is shown. This does not occur when running zoom-us from NixOS 21.11. A strace ends with the following output: Observe that the zoom process is killed by SIGILL shortly after the application window is shown. The Zoom application proceeds to run without crashing. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nI stepped through commits that touched zoom-us. The behavior appears to start with (no launch), resp. with (SIGILL observed).\nI can‘t reproduce this issue, at least I get into a video call and to making faces at myself for a minute without issues. I am running zoom 5.10.4, on x86_64-linux with kernel 5.17.9.\nI didn't see this on my system using the version from nixpkgs . It's expected that would not launch (it's fixed by the commits after). These should really have been one commit, or in another order to allow bisection, but oh well.\nFrom what I can muster, SIGILL is caught while in It looks like is part of the binary distribution of Zoom and not built as part of nixpkgs? But I figure if can run it with no crash it’s more likely an application bug rather than some ABI issue?\nI now notice that one Zoom invocation results in two coredumps. The other one is a SIGSYS arising from libc. I’m not really versed in such dark magics, but FWIW:\nI really don't expect to be able to debug this, but could you post the output of and (and maybe if the output is short, particularly around the marker) from gdb on the SIGILL core? To me this really feels like a zoom problem. You don't happen to be using an ancient or unusual processor do you?\nIt looks like version 5.10.6 (3192) is out for linux now, so it might be worth trying that too.\nCheers, that seems to resolve the problem. Not really, it’s an i9-10885H. In case you’re still curious for whatever reason, I’ve attached the information copied from gdb. Otherwise I’d say this issue is obsoleted by the new Zoom release. (NB just said “No function contains program counter for selected frame.”)\nmy zoom is dying too, I got it working for now with\nThanks, that's pretty interesting; something jumps into the middle of an instruction. Good that updating it works anyway."}
{"_id":"q-en-nixpkgs-8e9a4feb09931a46d86f699e79ec1c1adf6fc3ca8d5eac73dc2a9f0290f446fb","text":"ACME certificates cannot be provisioned automatically, even though run in manual mode succeeds. This seems to occur whenever allocating a new certificate, and possibly when renewing one. System: (Flounder) Nix version: nix-env (Nix) 1.11.2 Nixpkgs version: \"16.03.1168.\"\nHi guys; if any of you know what's going on here, it'd be greatly appreciated. From what I can tell, the Terms of Service are out of date for nixpkgs's version of letsencrypt - and this makes the service pretty much unusable. If you have any ideas about how we can fix this, I'll be happy to investigate further and make a fix.\nLuckily only happens when creating a new account, so renews are not affected. Thanks for reporting. Fixed. :beers:\nAwesome; thanks for the very fast fix! I'm wondering if there's a more permanent way to solve this - being unable to deploy new servers whenever letsencrypt releases a new TOS seems a bit fragile. Maybe we could add a configuration option to set the TOS hash on the command line when invoking simp_le?\nThat's a good idea. Also as we're right before releasing 16.09 we should add that option before the release so NixOS users can easily update the hash themselves if needed.\nHi! Even if the new configuration option to set the TOSH hash would only appear in (makes sense, it's a new feature), is the hash fix () going to reach ? It doesn't seem to be there for now:\nThanks for the reminder. I'll look at it. The only potential problem is that is at another version in 16.03. The patch should apply cleanly, though.\nany news? If someone need a fix, just add the following snippet to your configuration to override the package with a patched one: Then you can use the options as usual! But I hope the patch will soon be available in :)\nAh, thanks for testing that! In that case I'll simply backport that patch.\nOk, backported. Will be in the 16.03 channel soon. I'll keep this issue open because I want an to add an option for 16.09 to specify the TOS hash in the module.\nConfirmed to work on ! Thanks a lot!\nare you still planning on doing this in time for 16.09?\nwhat are actionables here for 16.09?\nWe won't implement this for now because we need to replace simp_le anyway.\nThis is happening again?\nThat fix worked for me. I've no idea if it breaks anything else for others."}
{"_id":"q-en-nixpkgs-99a445044bb65b6af5ab63154499802d44790221fd5837ee4b6e652fd40a9c01","text":"This is from nixos-unstable current. For some reason, I get build failures with ZSH. This is with a nixpkgs (system=x86_64-linux) and nixpkgs.crossSystem set to . This seems to be the same failure (I checked, its the same derivation): Steps to reproduce the behavior: ... ... ... A clear and concise description of what you expected to happen. If applicable, add screenshots to help explain your problem. Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\ncouldn't fix it. I'll leave this assigned to myself so i remember to take a another shot at it in ~6 months\nFixed by"}
{"_id":"q-en-nixpkgs-82ce395cd98a739455047839bd278ab9a2dfa156c39d4d20d05ee5e96de71669","text":"Project description Metadata homepage URL: source URL: license: Business Source License 1.1, change to Apache License 2.0 on June 1st 2027 platforms: linux\nCan you check out my PR if everything works?"}
{"_id":"q-en-nixpkgs-11f4a61c03fb3191b75c7ae6586a09e98434d826530ac1fc6b28b87c5371a2cc","text":"As of nixos/nixpkgs (i.e., today) seems to be failing due to an unsupported command line switch being passed (see ) Actually, starting doesn't fail, but some (all?) of the other command line arguments are ignored. For example, I tried setting , but see how insists on starting on port . Removing the switch makes start as expected, respecting the other given command line switches, but is this what we want to do? cc\nthis was broken due to Catalyst bump. See duplicated - I don't know how to best fix this as switch might have an impact. If someone has time to fix it, I'd be very happy :)"}
{"_id":"q-en-nixpkgs-84f60b87c319f9680005f54ba1812c2066102a53b00832d19ae7f74101e12d9f","text":"The systemd-boot image built in nixpkgs has gaps in the section table. This leads to signature verification failure with some UEFI Secure Boot implementations (e.g. Dell XPS 13 9310) Using NixOS 22.05 @ Using some old version from NixOS 21.11 that I still have in the store: N/A Allegedly something like this was fixed in systemd upstream and the commit is present in v250.4 but for some reason it still appears and Secure Boot fails. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. system: host os: multi-user?: sandbox: version: nixpkgs:\nRelevant:\nThis commit might be the culprit: I will try confirm this, but since it requires rebuilding systemd, it'll take a while.\nIt's more likely a problem of sbsigntools:\nDoes your hardware still correctly validate the signature with this patch applied?\nYes, signature verification works with this patch (and only with this patch). Hardware is Dell XPS 13 9310. See also"}
{"_id":"q-en-nixpkgs-2eebf1d8a8cae0ba27ca6e423ae9d3c99632ddf6585f0d366eb97ab3cfaf3a82","text":"Project description describe the project a little is a highly configurable dashboard for Neovim. It's an alternative to Metadata homepage URL: source URL: license: MIT platforms: Anything that can run Neovim, I would assume"}
{"_id":"q-en-nixpkgs-7026baa4b985f1b7bb817e65670864da1e168a8039d13239e84768ee85e4e118","text":"Project description GTK mc-like dual panel file browser ::: very nice for mc users... :) \"Double Commander is a free cross platform open source file manager with two panels side by side. It is inspired by Total Commander and features some new ideas.\" Metadata homepage URL: source URL: license: platforms: linux, ???...\nMan, this is a Pascal app! I never tried to build such a thing before...\ni'm just suggesting because i'm migrating from debian where i use it... :) ... though have no idea about building or packaging... :( debian has it, if it helps in any way.... i use doublecmd-gtk\nDebian rarely provides their build scripts. I am reading the ubiquitous AUR (). However, I have seen that the GTK they use is the old and abandoned-by-us GTK2.\ni do not \"depend\" on gnome/gtk... and was about to eventually give a try qt software... so feel free not to oblige to my request, though i do think is a nice piece of software that deserves a chance of availability... maybe wait for them to upgrade their dependencies... ( hour after ) ... i just tried krusader... and filled this bug report: O.o ... so seams there is not a nice working dual panel graphical browser available... :(\nCoincidentally, someone on the Discord server was asking about this application like a week ago and I wrote them a package expression to use. I don't have any interest in maintaining it myself, it sounds like you might be interested :wink: This was adapted from which has some modularity for GTK/Qt, in case you want to get that working as well.\nPerfect. Friday I will look at this.\n\" merged commit into NixOS:master 5 hours ago\" does this mean it is packaged?... how can i try?... i searched the unstable channel and cannot find doublecmd package.\nIt means that now the Hydra bot will build the caches in order to include this package. You can follow the progress of it in the link below: When it reaches nixpkgs-unstable you can possibly call it.\n... and is now available... :) ... and installed... :) Seams to work fine, except... When i try to \"Commands\" -\"Search\", \"New search instance...\" or \"Synchronize dirs...\", sometimes i get a \"Function not supported!\" error... :( ... After opening the program, if it happens, it always happens until i close it... If does do happen, then does not happen until i close it... o.O?... And now that i'm trying to repeat the error, all seams to work fine... Have no idea what was that... The debian version never gave this error... Since i'm not being able to repeat this error, i will not report it until if it ever happens in a repeatable way... (5 min later) Yap ::: no error for several program restarts, cannot repeat it, have no idea on what may have caused the error (maybe is just be from sharing my home (~) with both versions, of debian and /nix). & thank you for packaging this... :)\nfor mere curiosity ::: do you have any plans to also pack the \"qt\" version?\nIsn't this the QT version already? Or it has support for older QT releases?\nlolz ::: yes, you are right... sorry, it does look like gtk version, here in my debian, i presumed it was the same... so then, do you have any plans to also pack the \"gtk\" version?...\nI don't feel motivated to do it because it uses GTK2, an abandoned release of GTK.\nThe Free Pascal support for GTK3 is in very rough shape. GTK2 is preferred and Qt5 is getting there. what problem do you have with the qt5 version that gtk2 will solve?\ni have no \"problem\" \"with the qt5 version\"... :) ... it works just fine... :) ... unlike krusader ::: ... :( I just asked \"do you have any plans to also pack the \"gtk\" version?\" out of curiosity... :) And with the small concern that if some one only uses only gnome/gtk and installs the doublecmd (qt5) ::: will be also installing some qt5 libraries just for doublecmd, when the necessary libraries for the doublecmd (gtk) are probably already installed and being shared among other gtk programs. For example, if i want to make a small-as-possible live CD based on gnome, and the only installed qt5 package is doublecmd (qt5) ::: some space could be spared with the doublecmd (gtk) instead... O.o ... i presume. So, there is not a big \"problem\", that i don't mind now ::: but is also not perfect... i do undersand that might rather wait for \"Free Pascal support for GTK3\", and agree that adding \"GTK2, an abandoned release of GTK\" may just be a swim against the current and a waste of time (if no one needs it for a stronger reason).\nFortunately space is cheap, so it doesn’t seem like something truly worth optimizing for. Once you start dragging in graphical programs, you’re going to carry a lot of libraries. Use GNOME Commander in this case? I know it’s not packaged in nixpkgs, but that should be trivial to sort out.\n\"space is cheap\" ... o.O?... & what does life has to do with price?... ... and sure GNOME Commander could be a solution ::: but is not in nixpgks... (as you mentioned)... and is not double-commander... maybe is better, i have not tried it ::: is also not in debian stable, that i use... by the screenshots i would presume that has less features and looks not so nice. ... on my part, for now i'm happy with having dblcmd qt ::: i have not yet decided which to choose: qt, gtk, other or none, to cross with my xmonad... and still far from any kind of release, if ever any... and just scouting for the possibility of using nixos, or debian, or both.\nYour first argument fails in the point that, well, GNOME is mostly GTK3+ nowadays, whereas DC is GTK2. Between a well-maintained flavor with an extra library versus an abandoned extra library, the choice is obvious.\nIt has nothing to do with neither life nor price. The point is that the impact of bringing in an extra library (or set of libraries) in a desktop environment is negligible in most cases. I appreciate that there might be corner cases such as a bootable image where space is limited. Packaging it is most likely trivial - someone just has to want it bad enough. This is getting off-topic, but having used Debian both to run servers (50+ instances, so nothing big) and workstations, NixOS is so much nicer.\ni do agree with Anderson ::: \"the choice is obvious.\" ... and sorry, for bringing \"life\" into the subject... what i meant was ::: i don't care about \"cheap\" but rather well done... o.O?... a \"cheaper\" typical size DVD ::: costs more and takes more space than a mini-DVD, which does fit a pocked... & yes, \"This is getting off-topic\"... but the thread is closed, i presume that now we are just drinking coffee, right?... :) ... \"NixOS is so much nicer\"... o.O?... lolz ::: at consuming hdd space and having broken packages?... i like debian because is stable, and distinguishes between \"required,recommended and suggested\" dependencies... i use debian since 2000, and every time i try some other distro ::: i end up getting back to debian... :) ... debian can be small, can be big, is quite stable, and others are always missing something i need... :) ... as for nix ::: i do like the one-file ... but still find unbearable when i bump into a \"broken package\" (why is it listed for install at all?), and miss the suggested/recommnended variations of packages... NixOS is yet not nice enough to replace my debian... :) ... and debian + /nix is very nice enough for now... :) ... i just run 3 personal machines ::: \"so nothing big\" at all... :)\nThis isn't particularly constructive, but I'll bite. \"NixOS is so much nicer\" When you have multiple versions of the same package, you are bound to take up more space than on a distribution where only a single version can exist. The entire closure of bash (as an example) with all its transitive dependencies doesn't take up more space than bash and its dependencies on Debian. Complaining about NixOS using more space than traditional FHS distributions is a bit like complaining that water is wet. Sure, we have broken packages. All distributions have broken packages (for various definitions of \"broken\" of course). Certain types of packages (KDE applications on !NixOS) is less likely to be tested well and as a consequence will be more likely to break, that's correct. Do we have more or less broken packages that others as a percentage of all packages? Are they more or less broken? I can't give you a quantifiable answer. In my experience broken packages has not been an issue. Because the existence of a package in the repository is orthogonal to its build status. Exactly the same way that a package in debian can also be installable but not fit for purpose due to bugs (or conflicts or pollution from the global environment and so on). Apart from overriding a given package which is of course always an option, some packages will allow you to toggle various features, but having an application's runtime environment be a product of the global environment is not something we strive for. It's pretty much the antithesis of what we're trying to do here. That's fine, you do you. I was simply making the point that for my personal use-cases as well as those of the company in which I work have been a tremendous upgrade from Debian. Forget about \"not even the same ballpark\" - it's pretty much \"not even playing the same game\".\nO.o ... please do so... :) ... i think is constructive... i' learning about stuff... :) ... and ::: just having another coffee... :) Debian takes me much (much, much, than with recommended) less space, because i only install required packages, never the recommend, never ever the suggested... :) ... nixos seams to do that some extent with configuration and package options... :) ... but still feel much less clear and customizable to me... :( ... or maybe just don't know, NixOS yet enough, how easy that would be... o.O?... ... though never seen it on debian stable... since 2000... :D O.o ... lolz ::: & i just started now, trying to use NixOS... O.o ... & should i report this brokenness it as a bug?... (but \"thanks god\" ::: it worked on debian) ... seriously ::: please answer this... :) O.o ... lolz ::: am i \"experienced\" or what?... (just kidding) O.o ... never had packaging problems with debian since 2000... :) ... have you? ::: lolz, would be ironic... :D sorry ::: did not understand this. O.o ... are \"you\" not trying to please everyone and everything possible?... is not that what \"you\" also want?... O.o ... yes ::: i'll end up using both... :) ... i'm comparing with debian and sharing my transitional experience for \"you\" to be aware of it... :) ... i do find the nix approach very nice... :) ... and single file for the whole configuration is really nice... :) ... all my haskell work now goes into a nix-shell of my Debian /nix, that i test in a NixOS... o.O?... \"you\" do brag about good haskell support ... o.O?... and lolz ::: i bump haskell broken packages?... lolz... :D so ::: don't brag so much just yet... :) ... & make this \"bite\" ::: \"constructive\"... :) ... better yet ::: more construct-able... :) ... and more stable when announced (by being in the NixOS channel ) as stable... :) ... and ::: i hope the \"bite\" tasted nice... o.O?... i have not much more food for now... :( ... with any luck (for \"you\") i'll just keep bumping into bugs and brokennesses and report them... :)\nJust to be clear, the bokenness i mention above is also broken in NixOS, i did try both.\nA bit nore spacing concerns... I'm now upgrading NixOS... noticed all the -doc packages passing by... is there a way of switching them all off? ... and is there such thing like a central html index, linked to the -doc files installed? how can i access them?\nOh well, if you are comfortable using Debian Stable, your needs are such that NixOS won't be of use to you… NixOS is about avoiding conflicts between unrelated packages, supporting rollbacks so you can install fresh stuff with less risk, etc. If you want a minimal setup with Debian Stable, sure, that's another way of avoiding problems that Nix handles for large combinatations of fresh packages.\nO.o ... uau ::: not bad, for a dictator... o.O?... good thing i have a will of my own... :D but thank you, for the more clear answer to my question: O.o ... guess not... \"NixOS won't be of use to you\" ... :( sorry, i had not realized that NixOS just targets a specific audience ): ... as you explain, is not good as debian ::: to make a smallest live-cd, with no repetition of packages and with no -doc packages... :( ... ok ::: upon your advice i'll try to use debian for that... :) ... ty And please notice ::: that you just failed in answering any of the questions i made... :( ... which would actually be even more useful to go with the coffee... :) ... suggestion ::: \"get your self a cookie\"... :D\nI could even try to bite this, however my learning and attention problems trigger as a hell before such a badly written text. For someone complaining about elitism and bloat in an obscure and niche distro, your text does no favor being such an obfuscated and unreadable blob.\nSorry Anderson... I also was happier when we were just 2 making the more useful dialog already closed. Also my \"learning and attention problems trigger\"ed when someone else when off the good interpretation of my writing... O.o ... lolz ... have no idea of you are talking about. ps ::: according to you published location ::: Presumo que sejas brasileiro, e falas portugues... :) ... a minha lingua nativa ::: sempre escrevo melhor o portugues (sem acentos nem cedilhas porque o teclado nao os tem), que o ingles... e' isso ::: peco desculpa pela falta de qualidade do meu ingles nao nativo... sei e' escrever o meu codigo, e apenas aqui venho para me-vos tentar ajudar... seja ao debian, ao nixos ou whatever... :) ... mas se te estou a fazer confusao ::: ja' nao gosto... o.O?... e menos a ti ::: que empakotaste o dblcmd tao eficazmente... :) ... (lolz ::: maybe someday i'll consider inviting you for my yet non-existing crew) ... enquanto o extendido off-topic apenas esta' a deixar mais duvidas que respostas... :( ... portanto vou ficar por aqui ::: e evitar dar alento a \"elitismos\" maioritariamente incompreendentes... :) [ & bom dia para ti, obrigado por tudo, ] pode ser que nos encontremo-nos noutra thread. For the English speaking ::: if thou have no answers to my actual questions, please don't waste our time in replying ::: i came by to learn and help a bit, even if for now if is just by complaining and asking... :) ... sorry for my noticeable lack of \"social skills\" ::: thanks god for that... :) ... unfortunately written words don't carry the voice of the writer, nor attitude ::: and are quite prone to miss understanding... in future threads i'll try to hold my self from expressing so at ease. [ & have you all hopefully a nice day ]\nThis conversation has really veered off the original path. If you have any questions, I suggest taking them over to which is a far better forum for those types of conversations. I would however recommend considering if your writing style helps or hinders the chances of constructive feedback."}
{"_id":"q-en-nixpkgs-3175f034fa4c89202ca68f1b6a868cb38219da16cb930b3203ac54a1f8b8e9be","text":"With and , I still have to type my LUKS passphrase in qwerty. The logs say this: but not much else apparently. The vconsole is set up fine by the time I get a login prompt.\nDup of"}
{"_id":"q-en-nixpkgs-79087ce612c9924f9ddad5214dca59c702df537936c3b5b74618e39ad2d8d304","text":"Can anybody provide a pointer on why allowing dangling commits to be fetched is a security risk? I can't seem to find any details on this. Is the risk because people assume \"unreachable\" means \"deleted\"?\nThere's a guess in is the original author of the documentation (see )\nHrm, that guess has to do with github's PRs, not with anything in git.\nWell, I found a reason why git would want to prohibit this, but it isn't related to security. If you a branch, the commit that used to point to is still available in the repository -- at least until the next happens. If people are allowed to fetch commits by their hash even when those commits are unreachable, a force-push like this would put those commits at risk of being garbage collected. But nobody would notice the breakage until (potentially) much later, when a happened, and by then it will be too late to recover the lost commit. But since users are forced to mention a commit from which their desired ref is reachable, they will notice the breakage immediately. Most important of all, they will notice while there is probably still time to rescue the now-gc-able commit. I think this is the reason. If nobody else can come up with a security-related explanation I'm going to submit a PR to remove the word \"security\" in ."}
{"_id":"q-en-nixpkgs-84012a729ec08fb4dc06781f5db3df44563c781d56cf9a7f36802006b4a2dad0","text":"Steps to reproduce the behavior: Please run and paste the result.\nLooks like there is already a PR to address this at And fwiw I had to bisect this to fix my config, first bad commit is\nClosing since the linked PR was merged. (Not sure why merge of the linked PR didn't close this...)"}
{"_id":"q-en-nixpkgs-055a489bfb0124fd97eb9bca66e437900b90fd392f911e205c45e256c2ace099","text":"Project description vibrantLinux is a GUI tool to serves the same purpose as for Windows, which to modify the saturation of any given screen whenever certain programs are launched/are in focus. Metadata homepage URL: source URL: license: MIT platforms: Linux Only\nBump. I have some low vision issues and \"vibrance\" makes it much easier to read the screen. Thanks!"}
{"_id":"q-en-nixpkgs-042180150b9ce72435f424c15157bafaf10fb87201ff72ee275127cbd3b92d0f","text":"Steps to reproduce the behavior: Added 'kanboard' to Then ran My machine: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result."}
{"_id":"q-en-nixpkgs-fc835b86e67fed78ee4cf71834be1ece4ce2d1bb1db0df166dd2ffa159691a62","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: rustup <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 1.24.3 desired version: 1.25.1 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-92210dd6c925d90762bdec2f48ca9e5155d3d76885accf66892b16f1daa4dd67","text":"Linux kernel 4.6 on the stable channel cannot be compiled due to an outdated patch. The patch is removed from the master by the commit but it is not removed from the stable channel. Please backport the commit.\nCc"}
{"_id":"q-en-nixpkgs-f5af9dcc172c06a636a76b81b0634773ada43f482e3e35e3a60521026203fbc6","text":"The system-wide LDAP client configuration in is ignored by Openldap client programs. Steps to reproduce the behavior: settings into via or any other Openldap client Openldap client programs should obey the system-wide configuration file in . strace shows it's reading the example configuration from the package instead: Since commit the openldap package no longer passes to configure, it only passes to make instead. Please run and paste the result.\nIt seems a bit dodgy, but this might be a solution: plus corresponding changes to all the packages and tests that refer to . I'm just rebuilding a gazillion packages on my system from source, so I can check if that works properly...\nyou probably meant to ping and instead\nOops, sorry about that. I forgot that the maintainer handles listed in the package do not necessarily correspond to their Github usernames.\nI realised that changing is probably unwise. Not only does it break some other packages (like python-ldap) and some Nixos tests, it will also break any users who have configured slapd to refer to the schemas under , as I myself have: And changing is not necessary to fix the bug anyway. It seems that the approach of passing differently in compile and install works. I'll send a PR for it."}
{"_id":"q-en-nixpkgs-aaeea236773264a49dc232214f381632b4285374c0638b248b1a466d91462eb5","text":"The change to use at broke evaluation of The expression was changed to reference but that is undefined. It evaluates and builds with the following change? What I'm not sure about is if this is all that is required or if there is anything more before I submit a PR. System: (Flounder) Nix version: nix-env (Nix) 1.11.3 Nixpkgs version: (local git checkout with patches)\nAlso I'm not sure if any of the other expressions changed were similarly affected. Quickly testing the expression appeared to evaluate successfully despite not having as an input."}
{"_id":"q-en-nixpkgs-f7a1f96650ff9b62fae120a017abb475978e8410f8dd068f3dbad06d0bf9613a","text":"Per the , can only be installed on 32-bits systems. Is this still correct? It's a rather old expression, and given that other 32-bits software can be installed on 64-bits NixOS just fine, I'm not sure whether this needs to be changed.\nI think there was another attribute for as well, but I am not sure about that.\nthis should be documented actually\nOr the expression could be fixed to transparently install the 32-bit version. Like this does (I think):\nYou might be refering to there?\nI thought there was also something called without the part but I guess I am wrong with that. In any case, the solution proposes looks like a good one?\nDoes the package work actually? I'm trying to test it, but it seems to be downloading forever.\nOK, it seems to play fine."}
{"_id":"q-en-nixpkgs-31699079a7e4cf88355acfc36bb71c1436fc8ff7b3381803805f7f3dcfe9910c","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: minio <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 2022-05-08 desired version: 2022-07-15 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-cb97cefccf48f32f704d5f869e0e42ea6b09aa71c91fad296894e4354787cc0e","text":"When a user starts and opens a file dialog the application crashes with (Jameica 2.10.2:): GLib-GIO-ERROR **: 22:23:21.974: No GSettings schemas are installed on the system Steps to reproduce the behavior: jameica a \"workspace\" -manage backups -restore different backup Opens a gtk file dialog jameica crashes on every file dialog for me. the backup path is just the easiest accessible and something a user may come along quickly. i was already playing around a bit with the wrapGAppsHook, but i just dont get it running atm. so maybe someone else want to take a look on that. i am running a kde session on wayland, but i doubt that it make any difference on gnome or other DE. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\ncreated a pr for it. my first tries in adding wrapGappsHook did fail as i did not get, that the order for nativeBuildinputs do matter."}
{"_id":"q-en-nixpkgs-da4f9cc9f3729b4c932fff48dc81778c351f63537f56c939e048f87b7f8b8302","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: tutanota-desktop <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 3.95.4 desired version: 3.98.15 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package --maintainers: Version 3.95.4 is not supported by Tutao GmbH anymore. I don't know how that exactly works with Nixpkgs, but it would be desirable to update tutanota in the stable 22.05 channel too. Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-68f97dbb3d8a5a919595e742ba8c8e3c46834b5c03d54f00cac440fabad291c1","text":"name: flexoptix-app current version: 5.11.0 desired version: 5.12.1 maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-a91e32823e70cfe6d4e1b55d1beafc948ae6b2b219daacea9dc57ba4a1faf107","text":"Steps to reproduce the behavior: See . Regression of Blocks: Blocks: porting GSound and GNOME Clocks to macOS Notify upstream: libproxy/libproxy <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\nA possible fix:\nThat looks like an upstream bug – the build system should never use functions it is undefining. And either way, on Darwin, I would expect system JavaScriptCore to be used instead of Duktape."}
{"_id":"q-en-nixpkgs-6c8d9bcbdd13e6b8e50af77aac1a4d0876bf0fa6669382247b054feedef806d3","text":"Steps to reproduce the behavior: git-town on MacOS Please run and paste the result."}
{"_id":"q-en-nixpkgs-4b2cbcfacbd89b0774bccace43a1d7211b4462b2806bbb256083bde39fc305b4","text":"If I try to run Solaar via Rofi nothing happens, when trying it via terminal the following error occurred: If I open a nix-shell with and run solaar it opens and works without issues. Relevant configuration: Steps to reproduce the behavior: system from unstable to run solaar Solaar GUI main windows opens I've tried to find out more and was directed from that it could be related to missing but regardless where I it it didn't solve the issue. Please run and paste the result. nixos-unstable -\nsee\nThanks :)"}
{"_id":"q-en-nixpkgs-a4bcd25fa5b713fa03f700a379f48bcb68d6ea9974097f00b97aa323e1ca8e90","text":"Putting this out here in case anybody else is seeing this. On am Acer Swift SF314-43, the touchpad goes missing with kernel 5.19. Works again with kernel 5.18 and no other changes. Steps to reproduce the behavior: to kernel 5.19 that it's broken With kernel 5.18: With kernel 5.19:\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\nThis happended to me as well, on a Tuxedo Aura Gen 1 with an ElanTech clickpad and NixOS 22.05. Comparing the output of between kernel 5.18 and 5.19 I found three modules, which haven't been loaded on 5.19: , and . ~I them to , ran and after rebooting the touchpad is working again.~ Adding them to in my NixOS configuration did not help. Edit: Was on the wrong kernel version, when testing if manually adding the modules would help.\nAh, on 5.19: Kernel rebuilding time. Haven't done that in a while...\nSo that did it. I didn't need to specify the list of modules, only add the kernel option.\nThat fixed it on my Tuxedo as well."}
{"_id":"q-en-nixpkgs-2f461e8b78cd53278ce8593c9cd1891910ae3f17b6c0103a8b3ae494ed769416","text":"Right now, we have sections covering Perl, Python, Ruby, Go, and more. is another popular language and we have an easy/automated way to add new packages to nixpkgs. We should make this easier for people who don't already know how! I plan on tackling this based on instructions I received over at\ncc\nSome of the entries in the document have a comment like Should I enhance the in and then generate the XML file? Can anyone clue me in to how they're generated?\nThey are generated in (using pandoc).\nOnce I've it to , how can I actually generate the XML?\nThe XML is generated during the build. To have it in the manual you need to include it with in . The confusing part is that refers some files that exists only at build-time (the files generated by ). It is also possible to keep the build directory with to inspect the generated files."}
{"_id":"q-en-nixpkgs-7d23bca2afabe244e001ea1f5ff411bdf8a2bece327125c5e4df1a0119e16fcc","text":"glslc binary in shaderc has invalid signature Steps to reproduce the behavior: to compile shader on aarch64-darwin to SPIR-V should not be killed"}
{"_id":"q-en-nixpkgs-dd749394ca6f9ef8dec044c1f97ed1031c2283da667ca4fc89f0ac2387af4dc8","text":"Easyeffects crashes when Echo Canceller is used (at least when applied on input). I've been told that it works fine on arch linux. Though it might be my hardware that somehow makes it crash. Steps to reproduce the behavior: Echo Canceller module to the Input Easy Effects to one of recording apps It doesn't crash It doesn't write anything useful in the logs, just exits with code 137. Please run and paste the result.\nI can reproduce it but it is not a regular crash – there is not core dump. It appears to be killed. Weirdly when I run it with: The termination is not caught.\nFor me it reproduced with gdb: Maybe it's terminated when child process is crashed?\nhmm, it seems to be something wrong with speex when configured to use the fftw library... not sure what, but this doesn't crash:\nThe same also happens in Speech Processor effect now:\nUnfortunately this didn't seem to do the trick for me, it still crashes with the Echo Canceller.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\nI have the same problem. The thing is, used to work OK on my system. Then I tried the module, after which this issue appeared. However, now it's persisting. I undid the change, even tried booting into old generation, but is still getting killed with the echo canceller on. For reference, this is what I used to try to enable the echo cancel module (not sure if it ever worked):"}
{"_id":"q-en-nixpkgs-8975aa5250159fe5f6e08ff39cfabfabd478ac819ca4dfa560eb6fe4ae0bfd97","text":"When using as a root user the following error is produced: . This needs to be included, see = \"source ${}/share/zsh-\"; Please run and paste the result."}
{"_id":"q-en-nixpkgs-4282c34f0d9a0257061fb3ef2ce242e23b512dbf587d15b10181118645268527","text":"consists mostly of syscalls (including ). There are even Ruby-specific workarounds. C/C++ would be a better language for the task. This should also reduce closure size and startup time of FHS-sandboxed apps (Ruby needs to load itself and its libraries from disk).\nCan't we just use to setup the chroot?\nI'm not sure if it's capable of doing all things that we need for this, but it should be worth checking. Can it run a container as a process (not as a systemd service), and can it be used not by a superuser?\nneeds root privileges. btw has C code to set them up, the rest is probably bash only?\nD: A pity. Well, let's continue with the old plan then. Unfortunately we can't do much more in Bash than we already do because we need to do mounts in the same process we set the namespaces, in children necessary privileges are already gone. I'm not completely sure about this but I remember that when I tried to write this in Bash completely long time ago (using for setting up namespaces mounts and chroot were a problem.\nHow about using this one?\nbind mounts, sysfs, tmpfs and procfs should be\nI can confirm that these mounts are possible from using unprivileged lxc containers. But you cannot bind mount something from the host this way, because it needs to be done after . UPDATE: ok even that should be possible if the correct order is maintained: unshare -mount -chroot.\nOh, I should look at it! cc in case he's interested in this thread. I think that checks if the user is root -- that's why this doesn't work. This means that we need to call by ourselves... (I was going to hit \"Comment\" but you were faster ~_^) I expect that LXC containers do user maps such that you have UID 0 in new userns -- we, however, need to retain user uid.\nwhy do we have to retain user id?\nMultiple user apps use uid and username for their purposes -- we don't want to lie to them. I can't give a complete example (I remember that Steam breaks) but for example we have directories which apps can find based on uid.\nI have another suggestion: the Go programming language. Bear in mind that I'm generally not a super big fan of Go (in fact, while I use Go extensively at work, I generally use Haskell for my personal projects); however, I think that a higher level, simple (in terms of specification) language would be conducive to maintenance and further development. The binaries should be pretty small, and the closure should be pretty small, too. I might give that a go (no pun intended) and see what we think.\nSeveral possible issues: need native and then before any calls to pthreads (or you'll get or something close); I remember that Go has a runtime so there may be problems; need several syscalls which may not be readily available, namely , and others. In Ruby I import them from libc dynamically and redefine constants from C by hand, but doing this (especially redefining constants) is one of the reasons I want to move from Ruby in the first place. However these are only potential problems -- I don't know if they actually exist! I'm actually interested in how this turns out.\nYou cannot use clone() properly because of the language runtime. That's why docker uses a little C for that job.\nAh, you must be referring to this from : I'll need to give that a thorough reading before I understand why the runtime gets in the way (unless, perhaps, you have an explanation of why/how pthreads mucks things up).\nOh, I'm guessing that's probably stuff like this: Though, if there are other details, I'd love to hear them :smile:.\nIf you refer to my remark about fork and pthreads, it's actually 's limitation. You can look at how I workaround this in Ruby by calling imported instead of Ruby's builtin. You will know if you encounter it -- will just fail. I think the best way to find out is with an experiment. (notice that I don't know about Go runtime and haven't even used Go before, so, as I underline again, those might be non-issues)\nInteresting. I referred to the , but I don't see any mention of failure when used with pthreads. I don't doubt you; I'd just like to know how all this stuff works.\nNo problem even if these were doubts ~^. From man page:\nAh, dunno why I didn't try reading the man page -- thanks for pointing that out :wink:.\nAnybody done any more work on this? Also, this should probably be tagged with security, due to\na rewrite in a different language will not solve your grsecurity problems. I also do not see how the rewrite discussed here is related to security.\nAnything related to grsecurity is security-related. A rewrite into a compiled language means that a setuid wrapper can be created, which means I don't have to run shit as root.\nYou can drop your privileges via right after you enter the environment as -- it's less convenient but you get roughly what you want."}
{"_id":"q-en-nixpkgs-e8dc5dcb30f4262ed938efe3be477937624b78d25a3c08703824e75909560bc5","text":"Steps to reproduce the behavior: build -L 'nixpkgs#curlie' <details</detailsNot a go developer. I think the issue is the same as . That requires a fix from upstream; I don't know whether this is the same, though, and my attempts to patch the build (and thus demonstrate it needs an upstream fix) have been unsuccessful. I used with the following patch. This resulted in the following error. I tried vendoring the derivation and adding the call to (using ) but at this point I hit the time limit I'd set to investigate this. Please run and paste the result.\nYou need to update vendorSha256, too. IF you get it working please create a PR."}
{"_id":"q-en-nixpkgs-a5ce8108a9eb1ec970f91d1d1f62cddb62e7667bf4cdcbffb714b5e64d68e5e6","text":"In GNOME keyboard shortcut settings, the list of categories based on and on my own experience with GNOME on other distros should be: Launchers Navigation Screenshots Sound and Media System Typing Universal Access Windows However, only the following categories are available: Launchers Screenshots Sound and Media System Typing Universal Access So I'm missing Navigation and Windows. Install gnome package and use GNOME. Open keyboard settings (press Super, type \"keyboard\", choose Keyboard). See the list of categories on the left panel. System: (NixOS: , Ubuntu/Fedora: , ...)Nix version: (run )Nixpkgs version: (run )\nI am affected too. Moreover doesn't lock the screen.\nI got that, too. Seems like that might be .\nAnother possible source of not working (if you're using GDM):\nApparently the Control Center derivation only provides some keybindings: the others are part of other packages like GNOME Shell or GPaste. The Control Center loads the bindings , basically using to locate them. But even explicitly passing the path does not make the Windows section appear – though, strace confirms the file being read. Edit: Nevermind, it looks like the navigation and windows keybindings come from .\nHey, have you somehow managed to get it to work using ? (Not quite sure how to interpret your edit). I tried messing around with that but so far I’ve only been able to break :\nTry . The error you are seeing is caused by missing GNOME shell SCHEMAS, which will most likely be in the current variable.\nIt looks like the error I was seeing was related to fish shell not having setup properly and was failing even if I didn’t attempt to change anything. I switched to bash and that particular problem disappeared, however I am now getting a different error message: However, this schema seems to exist in mutter, so I sadly have no idea what is going on here:\nRight, for that you also need to add share/gsettings/mutter-3.28.0 or something like that.\nThanks a ton! I finally got it to work using that!\nI was able to get it to work using .\nThe mutter keybindings are not available in control center in NixOS when . Explicitly adding mutter to fixes it. Should mutter perhaps be to ?\nI think adding it to dependencies of g-c-c is more in line with Nix. We can do it after we merge 3.32."}
{"_id":"q-en-nixpkgs-5ea7c5b8568434184f062601f7488cba89992adc1c38d36d2ba96ecf108e03fe","text":"On current , (4.2.4) group voice chat will be stuck on \"connecting\" forever, while private group voice works. This issue exists when connecting from different countries, using different account, with or without proxy, so it seems not really a \"connection issue\". a voice chat in a private group. it stuck on \"connecting\" phase forever. Successfully starts or joins a voice chat. chat and live stream are broken. (one to one) voice chat works without issues. 's version of telegram works. (Its font is broken though) Please run and paste the result."}
{"_id":"q-en-nixpkgs-8982b9473e513a1ab33ffd8d8b8d6063a3da275a6b4c40468141ba3c2bd1603e","text":"Try building\nJust ran into the same problem when updating my system flake. For bisecting purposes, nixpkgs was my previous (working) configuration. EDIT: bisection in progress, have to do some large builds now:\nExcerpt from : which seems to be the change that breaks , so we probably need to file an upstream issue.\nAs a workaround, something like this should work:\nNote that the downgrade would undo some security fixes; see PR ... which is also why we backported this problem to 22.05.\nI see...I'll wait to re-enable qtile then\nDepending on how much value you place on the failing test, something like this worked for me: Modifying the package in a general overlay triggered huge rebuilds for me, so I've just replaced it inside my derivation, which is what pulls in in my system configuration.\nHow would you modify the package with a general overlay? Something like this didn't work for me:\nThis global overlay works for me, without annoying rebuilds: Note, I'm using and this is imported in . I don't know if you can use it without modifications otherwise.\nI have a similar setup, I have (minus the let bindings I have in my config): So I don't see the difference between our configs...\nSomehow, the overlay proposed by worked. This reaches about the limits of my Nix knowledge, but maybe is required instead of simply ? Is the only difference I can notice."}
{"_id":"q-en-nixpkgs-8795d6d32e58f9507757b4816eaf917f27f7e8966a10da5892f6750c108b6fac","text":"When one installs mastodon via the following minimal config: Then there is no obvious way to run tootctl. or any other login user doesn't have the executable by default, nor can this be enabled in an obvious way. is not a login user, so can't (and shouldn't?) log in as and execute . Steps to reproduce the behavior: mastodon as per minimal config above ... nope. ... nope. ... nope. ... nope. Any of: a setting such that is installed for that user an explanation in the docs what to do as a login user (probably bad idea?) Workaround: CC\nI think upstream intends only to be run by the mastodon user. My way for doing that is just running and continuing in that shell. The reason why doesn't work is probably because sudo inherits PATH from the user you are coming from? I'm not sure I understand the concept of a login user enough. How would mastodon become a login user and why would that be a bad idea?\nIn the mastodon docs, it is described the following way: But is not possible, it gives the error They do recommend (which is what I meant by it shouldn't be a login user), but still they seem to do something different than nixos does. Maybe that's fine, and it would be helpful to document this somewhere, so people don't spend a long time trying to find it out.\nThis is default for almost all service users because they have no login shell. Use ` This is wrong. Sudo resets all path like variables like PATH and PYTHON PATH. Also inheriting needs to be turned on explicitly and PATH mist always ve inherited explicitly. Other services like nextcloud include a wrapper that sudos to the right user. This should be for mastodon, too.\nSounds like a good solution. Then we could make the command globally available.\nThat isn't? I would have expected that it is very similar to nextcloud-occ.\nNo currently it is only available to the mastodon user, so tootctl never gets called as the wrong user.\nThere is a hack in the nixos nextcloud-occ script that automatically sudoes if called by a different user. This could be adapted to tootctl I think.\nYes, this should be adapted. Changing the mastodon to a login user would not be the correct fix."}
{"_id":"q-en-nixpkgs-9f348d988efbd12c01214e8e8eeab863226e4a490317b78b7bc54f8b4097c506","text":"When exporting the graph, logseq throws a permission exception. Steps to reproduce the behavior: a new graph a few pages them public ()) the graph The graph should successfully export and opening it should show a read-only view of the graph. The logseq documentation itself is an example thereof. If applicable, add screenshots to help explain your problem. Bug was first reported in the original repo When running the AppImage with it's fine because it's extracted into , but when the app is in then it won't work because of the read-only status. I tested the flatpak version because I suspected the same problem would occur, but the directories are mounted read-write in a read-only volume. So, 2 options exist as I see it: fix it upstream, but I don't know Clojure nor Electron start the AppImage with instead of extracting it Please run and paste the result."}
{"_id":"q-en-nixpkgs-6b27e99607c9b09248b1e172673e899c8a590a0c0624774177a302b77cb91910","text":"On my M1 Mac, with homebrew-installed libvirt / qemu-system-aarch64, I'm able to run the VM defined in the config below without issue. However with libvirt / qemu installed via nix, I get: Some possibly related discussion . Looking at the , they're explicitly setting ; adding this seems to resolve the issue: I'm not sure whether disabling those failing tests is a reasonable thing to do. Steps to reproduce the behavior: to in A clear and concise description of what you expected to happen. Should run without , like it does with homebrew. Please run and paste the result.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\nFWIW, I also encountered this issue, and have to use the following workaround: And -ing the efi firmware & vars manually. Fixing this bug can lead to great user experience: virt-manager can work out of box.\nWhere did you get ?"}
{"_id":"q-en-nixpkgs-482fd8673585fe7c98847de5ccf84a66b96737eaaa047ef4db0c5870bd3fde7c","text":"To reproduce, here is an example: The related code is at Some nice to have: As suggested  should be something that can be accessible by everything and we are likely going to put our \"\" there and next we are going to import in basically every file. Adding to in is probably not a good idea since this is a desktop shell package you will end up having this variable everywhere, I imagine a treewide will be likely needed."}
{"_id":"q-en-nixpkgs-e4f4300eb1bd80ecd21fc0af2430b5787ac57f6d458483f50be3764ab76afc1e","text":"When running , I get Restarting the system did not help. Steps to reproduce the behavior: The program should launch without crashing. If applicable, add screenshots to help explain your problem.\nHello Thanks for reporting this issue, and sorry for breaking the nvtop package! I made a pull-request to fix nvtop: What do you think about this fix?\nThank you for your fast reaction ! The PR looks good to me. I hope it will get merged soon :)"}
{"_id":"q-en-nixpkgs-712d5b7f44df19162afcd5dbafa46ae6859ef7eb00cd5ca5240b0b590a30a6d3","text":"pam_p11 causes login segfault in NixOS 22.11. GDM and Gnome lock screen are also affected. Steps to reproduce the behavior: to log in on the console or run `: Bug was not present in 22.05.\nReverting does not help.\nFixed by: Can someone with commit access apply this patch?\nCan you open a pull request with that patch?\nI hope that NixOS can be unbureaucratic enough that PRs are not required for this kind of patch.\nIt's generally preferred to do basically everything through PRs, so some basic CI checks can run & maintainers can be pinged automatically. We can just self-merge though. I created containing your diff and adding a comment for explanation.\nThis seems to be whats causing my 22.11-beta19 install on an Alder Lake 12700H (no gpu) to not boot into GDM. This is a pretty heinous issue that should be tested for. The graphical installer doesn't install Vim, WPA_Supplicant or anything useful for figuring out what went wrong, despite having Vim and WPA Supplicant available on the installation media/calamares.\nWell I just did - you're welcome. That's what beta releases are for, aren't they?"}
{"_id":"q-en-nixpkgs-07a7f4dbe892227e938d560ea16281033c880bb2019ece41fff50de72a9b0d43","text":"i3 has been updated to 4.13. I attempted to update our derivation but they switched to and I don't know how that works :) If somebody could update i3, I would really appreciate that. /cc maintainers\nFYI, for to work it is sufficient to add to the (see ).\nThanks :) I think with the main i3 the issue was that the needed to change and I did not know what to set them to.\nYes, and indeed the and are already included with the source distribution for , but I had to add the for ."}
{"_id":"q-en-nixpkgs-3635650f75ebdbc7ddef5bbe27b00a995720d03117310c288c5b3ee1cc5ef93a","text":"The srb2 assets are outdated. Steps to reproduce the behaviour: the game or run the command while in game The that the assets are being downloaded from hasn't been updated since 2020\nI just noticed this pr, but it seems to have gone dead 3 months ago\nclosing as I'm making a new issue"}
{"_id":"q-en-nixpkgs-8134cc91aa1b30e156be8000d7086d595d40d47edc73e2d6c49d1c33dc37622b","text":"Specifically systems with AVX but without BMI2 Same as this debian bug: They solved it by pulling the latest stable 2.35 patches: ;a=history;f=sysdeps/x86_64;hb=refs/heads/release/2.35/master For NixOS, one workaround is to install a NixOS system to a usb drive with microcode updates enabled, and use that for the installation. This is because for this series of CPUs (Haswell, 3rd/4th gen) the BMI2 instructions are via a microcode update. Rebuilding the iso with this overlay also works, but this is a mass-rebuild: The diff is produced like this: Where corresponds to 2.35-220 and corresponds to 2.35-163 (our previous state).\nIt was attempted to update glibc to the latest stable patches in , but this led to some issues.\nA temporary fix without rebuilds would be via , I suppose. EDIT: or 22.05 release in the meantime, as I expect it shouldn't be affected and it's still supported for a few weeks. Real fix (taking more time, etc.): most likely update glibc to latest 2.35.x and somehow resolve the regression (and possibly others), and then backport to 22.11 as well.\nRedoing update as PR for now.\nThe glibc update is contained in and channels now."}
{"_id":"q-en-nixpkgs-3c59237d2f55e457fecdf94b553c79f1097e62ee8e9b175bf9080384bc19675c","text":"Running Dwarf Fortress () on darwin-aarch64 opens a window that is completely white. It receives inputs and sound also plays, but nothing is displayed. This might be due to the package not having an app bundle as an output (e.g. 'Applications/Dwarf Fortress' next to bin) Steps to reproduce the behavior: dwarf-fortress on macOS Window opens with graphics displayed ! <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\nI should have tested the files from the official website first, since it turns out it has the same issue. Changing PRINT_MODE to STANDARD made it display correctly. I don't know if the should still be patched in nixpkgs to include this so that you can leave the link to the store for the file. Feel free to close this otherwise."}
{"_id":"q-en-nixpkgs-b85f99a5ae9fc896cccac42a4f59880fe9b444ddaafcfaacc8c0ccc0a4ced755","text":"The postStart script of elasticsearch fails when a special values like or ( ) is used for services.elasticsearch.listenAddress as the script expects an IP or a hostname. Using an ip as listenAddress ist not always a viable option especially when generating NixOS images upfront for a dynamic deployment (e.g. on Azure or EC2) I am not sure how to address this issue. Perhaps one shall check netstat output to find out which ip was bound to? /cc\nYes i agree, we should implement another option for es url, I'm fine even with removing postStart. You can create pull request if you need fix asoap, I will put it on to-do."}
{"_id":"q-en-nixpkgs-e6ac1eae178d600b44695bd36329f01ad76705a47e805a0579e0f33a05d7fb21","text":"Can't launch discordcanary with NIXOSOZONEWL=1 Steps to reproduce the behavior: NIXOSOZONEWL = \"1\"; in environment.variables discordcanary with Discord should open without any errors I got this error Please run and paste the result.\nLibrary should be from\nPlease try with"}
{"_id":"q-en-nixpkgs-fc8771c7a80ea1109f02b54e2e024e507e8ccd8b7a27382ee06b9ec04f982120","text":"In the pre-start script for , we symlink the template from the nix store into : However, if changes are made to CUPS via the GUI -- for example, enabling network access -- this symlink is overwritten with a plain file. And it contains a line with paths into the nix store. This effectively freezes some version of the cups binaries into this config file, which is no longer managed by nix. As a result, CUPS can break, e.g. when deletes the referenced store paths. For me, this meant nothing would print, and I started seeing this error in : Manually editing to update the store paths on the line (while the systemd service was stopped) got printing working again for me. I guess we could make the pre-start script do this (with a sed or something), but that seems kind of hacky ... cc commit seems especially relevant here. System: Nix version: Nixpkgs version:\nHm, maybe the right approach is to make just put in the template, and then have it also ensure that points to the correct location in the nix store? Then CUPS can write whatever other changes it wants to its config file, while the nix expression continues to have control over the \"final destination\" of the path ... ?\nI think we want Nix to take priority here because that's what would be expected of declarative configuration and what's already done in other places. So as a quick fix towards this we can force replacement by symlinks. This should also be documented. However, we can leave some option (like currently exists for users) that reverses this and instead makes just insert and other required options. The best of both worlds, of course, would be to split the configuration into generated by Nix and values changed in CUPS on top but I doubt this can be implemented. Thoughts?\nFor the record, I had the same problem in OctoPrint service. It was solved by writing a small piece of software for updating YAML trees (like in Nix) and making \"patch\" current configuration with Nix-overridden values. Works perfectly but it would be much more difficult in CUPS ( is some kind of ad-hoc mix between XML and custom plain format).\nYeah, I don't see any directive on the manpage for . :-( Overwriting the config file on every restart of might lead to confusion if folks are trying to use the GUI to get things configured the way they want. In particular, when I made changes in the GUI to enable remote access, it modified some of the stanzas that're currently hard-coded in at So I'm not sure how I would preserve that under this approach. (I tried to just set via , but that wasn't enough -- I don't really understand why ...)\nIts man page says: ... but a bit of googling didn't turn up any pre-existing programs designed to read / write that format. (Maybe there is something buried in apache , who knows?)\nThere is But then you need to write Perl, which might not be an improvement."}
{"_id":"q-en-nixpkgs-da301ac6e280eb22927916e6cc6399556d01d9bbcd8bb4ffc4d6032e61e0b9c0","text":"<!-- Note that these are hard requirements --<!-- You can use the \"Go to file\" functionality on GitHub to find the package Then you can go to the history for this package Find the latest \"packagename: oldversion -newversion\" commit The \"newversion\" is the current version of the package --- <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- name: openshot-qt <!-- The current version can be found easily with the same process as above for checking the master branch If an open PR is present for the package, take this version as the current one and link to the PR --current version: 2.6.1 desired version: 3.0.0 <!-- Search your package here: If no maintainer is listed for your package, tag the person that last updated the package -- Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-307dcf2ac52123ea9f884b958e76d84279db1c6c069de6649bd6b193bf18b2cd","text":"explanation pnpm version is a little bit old. May I ask for update in unstable channel to v7.24.3? Project description Fast, disk space efficient package manager: Fast. Up to 2x faster than the alternatives (see ). Efficient. Files inside node_modules are linked from a single content-addressable storage. . Strict. A package can access only dependencies that are specified in its Deterministic. Has a lockfile called pnpm- Works as a version manager. See . Works everywhere. Supports Windows, Linux, and macOS. Battle-tested. Used in production by teams of since 2016. . Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin"}
{"_id":"q-en-nixpkgs-94f1a0f14b5fe791c25935d91d4308b6fa5ae668a50b7838300aa7e2af7e43f7","text":"Project description A cursor theme based in the capitaine cursors but themed with gruvbox and nord. Metadata homepage URL: source license: lgpl3 platforms: linux since it's based in capitaine cursors that use lgpl3 i think it's gpl3 too."}
{"_id":"q-en-nixpkgs-123829393930069415d9904447800401a8c966db2b48960252e152b20cb4681a","text":"The version of pillow rmfuse uses is vulnerable to CVE-2022-.\nSince rmfuse is packaged using poetry2nix you are basically the only one willing to update it or its dependencies. That's problematic given that you don't do so even two weeks after being notified of a vulnerability.\nI don't think interacting in this way is very constructive. There are different points of view regarding packaging Python packages, that's it. That an issue like this does not get picked up is not strange, people are typically busy. If you like to I suppose you could add the CVE to the expression. Whether the CVE can be considered a security issue in case of rmfuse should be considered as well.\nWhy not? I'm just articulating a problem with the way rmfuse is packaged right now that is obviously an actual problem. It's not technical though, just organizational."}
{"_id":"q-en-nixpkgs-d85c17bcf294aae24d18438ee3c997d49b763e2ab8f4c8441940413472f32ccd","text":"Steps to reproduce the behavior: in nixos-unstable -note build will failed Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\nYou can override paper-note by adding the following section and it should work fine."}
{"_id":"q-en-nixpkgs-420f7ff22085cd5f3d2835cacfdd9aa6c6ffda67c99e2795ccb6ecc3c921e9d6","text":"Project description It would be nice to have libreoffice-still-qt, to have a stable version of libreoffice (still) with kde integration (qt). In my undestranding, so far we only have libreoffice-qt which is the fresh version cf."}
{"_id":"q-en-nixpkgs-b518789d8e05f4123c21d20ea32c53a930711a8fea883605b774dff80ef1213e","text":"Project description Metadata homepage URL: source URL: license: SIL Open Font License 1.1 platforms: all"}
{"_id":"q-en-nixpkgs-eb8e4d88d485b00268633bd8cefc89c50e1e253e292a5e0f62d8c4d29c37ab2b","text":"gamescope crashes with this error when run from within steam's fhsenv: This check in wlroots causes that error because inside steam's fhsenv the /tmp/.X11-unix directory is owned by nobody:nogroup. I have tried simply removing the check from wlroots but in that case gamescope simply hangs. (Note that it's gamescope that depends on wlroots so this bug happens under x11 window managers too.) Steps to reproduce the behavior: steam to the launch options of any game the game\nThis seems to be a very old issue with NixOS\nwould allow to run with without having to manually chown /tmp/.X11-unix but running from within steam as described above is still broken. gamescope dies with this error: However : So that symbol is actually defined but Xwayland thinks it's not for some reason.\nI've encountered this issue too! I know got gamescope to run on NixOS - maybe we're missing something? See my comment here: Related:\nHi, I just wanted to real quickly add some context with my knowledge of how Steam works. this part is not the issue and is an \"\"issue\"\" about the steam overlay, it happens because there is both a 32 bit and 64 bit version of the overlay both of which are loaded when you launch a 32bit or 64bit game, so this is not related. It happens every time you launch a game on any distro without fail since a game cannot both be 32 and 64 bit. ERROR: object '' from LDPRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored. Everything else in this log seems normal, outside of this part: which could point to a missing library. the mentioned \"\" might be what we need to look at here. Also, thank you so much to for posting the workaround too, I was getting a headache dealing with this.\nHi, i wanted to ask if someone is working on this issue?\nIt's an open-source project, so unless someone with the necessary knowledge is interested in the issue (could be yourself), then you'll probably have to wait.\nI ran into this same issue before. The only thing that let me run gamescope through steam so far has been the solution from , but I am not sure why it works.\nI'm trying to add the patch in in , eagering for testing from you.\nThere seem to be multiple issues conflated here. The original issue with X11 socket ownership has been addressed. If there's another issue that prevents running gamescope in Steam, please report it separately."}
{"_id":"q-en-nixpkgs-4b8e9c09eea96c3f17be7f81bd6809ebe7534e3b3679c7a18ccce26db561916a","text":"Steps to reproduce the behavior: to build from source Please run and paste the result.\ncc maintainers\nCan you check if is working? Looks very close to the regression fixed in (did not get to yet).\nis there a cache for that branch? using with that started building the world\ndoes have an incomplete cache. If you are accessing at inopportunate time you might catch a few rebuilds. You can pick a specific commit from with most builds done. Most recent it yesterday's\nthat commit did not fix the issue\nedit: scratch that, tried a fresh build. works now\nClosing since it sounds like this was fixed by the merge of . Happy to reopen if I'm mistaken."}
{"_id":"q-en-nixpkgs-ffab4bddde8e86d487d9239bf1102cb595f5aee13352dbc6dd5ca587aeec14c8","text":"Using with seems to be broken with and . It fails with . also notes this. Running yields . The contains just an empty and I am using this flake's devShell: The same also occurs when using a shell with only and when trying to use inside a nix build. I am pretty sure that this is caused by in containing instead of (which it does for ). I don't know yet, what the root cause for that is or how to fix that. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- (So it shows up in the tracking issue )\nApologies for the delay in getting to this. Minimizing your example ultimately leaves me with: Staring at the output of ^ with reveals that is a linker script that appears to be passing in : For LLVM 14 this file looks like: But for LLVM 15: EDIT: whoops, my bad; just noticed that you already mentioned this in your post, I totally missed it Looking through the source, this file appears to be generated ; specifically the should be coming from based on . For Linux/macOS (i.e. platforms using libc++-abi) that should be set . Poking around a bit suggests that this is a bug in upstream; is not an interface library so it has no (though it has an ) and its is never set. And sure enough, looking at upstream reveals that has been (see ).\nApplying that patch and rebuilding locally seems to fix the issue; I'll open a PR (and add regression tests for this) shortly.\nThanks for looking into this! Building from the PR-branch now no longer fails in the way I described and the example works for me again. I have now found a new issue, tho: Importing fails with: For example, this causes this error (using the same flake as above, just built from the PR branch). I was able to work around this, by adding to the end of the phase of locally, but I am not sure if that's the right approach here.\n(Re I have a proposed fix for the _cxxabiconfig.h issue on llvmPackages_16: -- currently I'm inserting an additional include search dir for it. I also don't know if this is the best approach or alternatively if that header (and friends?) should be copied into libcxx or similar.\nI ran into the same missing header issue compiling using on . Maybe the fix isn't enough, and should be copied?\nthis change is missing from 15\nrunning into the same problem compiling some static libraries in nixpkgs using pkgs.llvmPackages_16.libcxxStdenv. Does this warrant its own issue?\nYes, file a new issue. Please also provide reproducibility bits, including the error message you see, the package expression you're trying to build, and the nixpkgs revision. I'm guessing the issue you're seeing is failing to build, which is a different problem specific to static builds.\n(If my guess is accurate, the problem you're seeing is a variant of ).\nsorry for bothering you, turns out it wasn't really this issue. Using my limited knowledge on the matter from what I understand, is not installed to like . Now, turns out (and ) is a default system header path (for CXX of course) and so is BUT is not so if one were to they would just run into since that is nowhere to be found. There are two solutions to this is as a system header path like we simply just install everything under on non darwin systems as well. I actually do not know why these are all not copied on non darwin systems. Please correct me if wrong. I might be completely wrong.\ncheck\nNo bother. I'm sticking my oar in because I'm interested in getting to the bottom of issues exactly like this. It's difficult to fix things it without precise reproduction steps, though. What you're describing there sounds very much like something I tried to address in (see ), which is on nixos-unstable. So it sounds like you may have found further issues, but I can't do much to help without a reproducer. In any case what you're describing sounds like a different issue to the original (which was ), so I suggest posting a new issue pretty much as you described but with a reproducer I can run at the terminal, since it is meant to be in the system include path already (per the link above). If it's not, that sounds like a bug worth filing.\nOkay I found the problem, and the \"obvious\" solution as well, you can find it here"}
{"_id":"q-en-nixpkgs-ba8787283d285391f1ed93baea2786eafc7bdb14415750a40530b22f0e6c2169","text":"Build of failed on x86_64-linux as of . This is currently breaking . python39Packages.tensorflow-datasets maintainers cc: maintainers cc: Other cc: n/a nixpkgs at commit This issue was automatically generated by . that created this issue. Internal tags: - -\nhuh this is a weird failure to me since fasteners==0.18 at the time of this commit which should be working... building locally on this commit i get a different failure: very peculiar..."}
{"_id":"q-en-nixpkgs-09fc1e8d006aa8c26eb33d1f94d85b1f9ed2c533fc5ffd924fcb8642a592aa73","text":"Currently, most functions work without problem. However, when I try to add new paths to search dictionaries, it crashes with the following error message: Steps to reproduce the behavior: goldendict Edit Dictionaries Source Add crashes A dialog should pop up to select the path. Reproduced using the latest master branch of this repo. Using startx and qtile as the window manager I've submitted a PR to fix this issue (see ) <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --"}
{"_id":"q-en-nixpkgs-c93d58690c6504f2e5ec33fc555c1a909d79c537e0ca021a7f6a816b088b4540","text":"sakura terminal emulator terminates when trying to change color by context menu with following error: GLib-GIO-ERROR *: No GSettings schemas are installed on the system [1] 9932 trace trap sakura run sakura, right click to popup context menu : options -select colors -(click on color field to open color chooser) Here is a thread that discusses the issue. it seems that the color-chooser gui description is missing that is needed by sakura. should include gsettings as runtime dependency. actually it has gtk3 package as dependency which provides those schemata, however it obvious doesn't look in the right path. this kind of error got fixed for System: (17.03pre98008. (Gorilla)) Nix version: (nix-env (Nix) 1.11.4) Nixpkgs version: (\"17.03pre98008.\")"}
{"_id":"q-en-nixpkgs-48fc381f275fcf61bb2cc2b2ddd0d2b95fa33ff049ed5169a537b8be7902a126","text":"Steps to reproduce the behavior: package is marked as broken I marked this package as broken starting with linux 6.2 because enables by default. The binary blobs included with the nvidia driver are not compiled with support for IBT and nvidia hasn't been able to update their internal toolchain to support it at this point. Rather than leave it broken, we should probably check if the driver is built using the proprietary blobs and add the kernel parameter starting with linux 6.2. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result."}
{"_id":"q-en-nixpkgs-1e242a274dcce00f500c16ed937e90e5f90699a75c3a398861064d73f5f09191","text":"A ends with the following: commit -rebuild switch System: NixOS: Nix version:nix-env (Nix) 1.11.4 Nixpkgs version:\nDo you have grub on an encrypted partition and have ?\nno, I don't have that option specified anywhere at all. Also, the error has since disappeared after moving to a fresher .\nGRUBENABLECRYPTODISK=y' in file That seemed to say that, no idea what the issue was then, so I'm closing this, please reopen if it occurs again."}
{"_id":"q-en-nixpkgs-94a9aee2ec28de03d173583e8274c7bce78f196330b4c0cb40560faf1af9b919","text":"The script does not (consistently) follow symlinks in the parameter, which it uses to locate the system closure to activate. This breaks when used with , which creates a symlink. When the kernel command line has , the script looks for , which does not exist. ~(Workaround: generationsDir users can pass instead.)~ Update: That doesn't work either. with will fail in strange ways, ultimately stemming from not calling . System should boot successfully when using generationsDir and systemd-initrd together. Symlink resolution was partially addressed in , but only for cases where is a symlink. Booting also fails if the given is a symlink, or if any part of the path to is an absolute symlink (because filesystems are mounted at , not ). I took a stab at fixing this in (), but it does not handle absolute symlinks correctly. Since generationsDir creates absolute symlinks to , this fix alone does not fix compatibility. ( changes those symlinks to relative, and together the PRs fix the compatibility issue.) Ideally, the script should follow symlinks in any part of the param, even if they cross filesystems. One option is to resolve the links inside the chroot. I couldn't figure out a clean way to make this work. (Can we trust that will have the we're looking for? We can't use because we're still trying to establish which system that is.) Another option is to somehow . (At time of writing, this question has no answers.) (guessing based on git log):"}
{"_id":"q-en-nixpkgs-73a59a1a4a0413995185dceb84efc23042436778118984a011b48cdd8a503fe8","text":"Package name: wpgtk Latest released version: 6.5.7 <!-- Search your package here: --- Current version on the unstable channel: 6.5.5 Current version on the stable/release channel: 6.5.5 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-b74b871a457692f97e400c80c0c41d611b521036afe465c242706ff607eb0c8d","text":"Changes introduced in forcing package (or whatever the option is set to) installation even if the set to . This is because the module is unconditionally adding the package to the : Steps to reproduce the behavior: both and to in order to avoid enabling the service or systemd overriding the package with . installed packages, for example using the command. An package is installed into the system: An package should not be installed into the system. I am on a quest to build a minimal NixOS system for usage in LXC containers and trying to trim things down to the minimum, without having to resort to forking, overriding, patching and maintaining individual modules. I am happy to open up a PR to move the package assignment to the block: but would like to check if anyone has any objections or reasons why I shouldn't be doing so. appreciate any insights ‍ Please run and paste the result.\nAs the author of that PR I have no objections to restoring the old behaviour."}
{"_id":"q-en-nixpkgs-e44c76875241fcb2f9013ddbdfb46c80188d24a1500b5b3e5070b66b5dfc6258","text":"Steps to reproduce the behavior: (the current version in ) I was having the same issue that was fixed by until a few days ago and built and worked flawlessly. I am currently using the version it is supposedly trying to build. returns , so I do not understand why the build is needed in the first place. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- (because you made the last commit) Please run and paste the result.\nI tried rebuilding today after the fix you committed (thank you for being so quick), I am still receiving the same error. Is there something in particular I need to do for the fix to propogate? I know that there is a little lag before a commit comes in from , but I thought 3 days later would be enough?\nThe fix is now on and , so if you are on either, I guess you just need to update the channel. Otherwise, the following command should work:\nFWIW, is also a handy tool for checking the status of a PR's movement into the various channels.\nthat worked, thank you. this is also very useful, will put it in my bookmarks!"}
{"_id":"q-en-nixpkgs-ff2b156278e5fb77404f10f235ae803b85f11e1392f2c353f0ac4c154aee300c","text":"Steps to reproduce the behavior: -shell -p quickbms Please run and paste the result.\nIt's possible that this package just isn't buildable One of the errors encountered is: is not supported in recent openssl versions."}
{"_id":"q-en-nixpkgs-187ae21c4060318706f76f649d88dae489d2ab3ec60eb2e9e116fb3673468943","text":"Project description CLI for interacting with Scalr. Metadata homepage URL: source URL: license: apache 2 platforms: windows, linux, darwin"}
{"_id":"q-en-nixpkgs-c2209d6bba8b3787d141e3b15de039117f31039d7ae6e2c3bdc944950ac1fc06","text":"Package name: oracle-instantclient Latest released version:21.9.0.0.0 <!-- Search your package here: --- Current version on the unstable channel:19.16.0.0.0 Current version on the stable/release channel:19.16.0.0.0 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! -- looks like it tried to upgrade v19 of instant client, but it is stale and not targeting v21. Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR.\nIt looks like you were the last one to touch instant client.\nPackage sources:"}
{"_id":"q-en-nixpkgs-e7e0e7619dc99cdebbb0cfa06debae0c8d35988659595163621e16713b4db823","text":"Package name: swayws Latest released version: 1.2.0 <!-- Search your package here: --- Current version on the unstable channel: 2022-03-10 Current version on the stable/release channel: 2022-03-10 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-87242f589026b73459dc7239055a6968b66182ba8e3c79a19bc0d56a085867ad","text":"This is a continuation of the topic from the initial \"upgrade datadog-agent to v7\" ticket - see for details on that. The is a metrics and log collection agent intended to run on linux systems for use with 's paid service. On other linux distributions, installing the datadog-agent includes the integration/plugin code, there is no separate install for those, or post-install steps to install the plugins (only config to activate/configure them). Configure the in , and add some extra integration, such as the docker integration. and look at the installed, the plugins are not available. You should see logs from the agent about those. When configuring the datadog-agent, the standard integrations/plugins (docker, nginx, apache, postgres, etc, etc) should be installed and available to the agent if the agent is configured to use them. Please run and paste the result.\nThere's still something wrong with the Python environment.\nMost likely we need to use pyproject to build it properly: That results into missing ddtrace package, which needs to be .\nLooks like setuptools wasn't building the and modules. and should fix the integrations. Datadog is planning on dropping setuptools, so we should switch over to pyproject. The blocker is that there are a number of missing dependencies that aren't on nixpkgs yet, but I've started working on that. the Docker integration was rewritten in Go and is now built together with the . You can enable it with: I have no idea why they're now called Yeah, we need to rethink how these integrations are enabled. Now that some of them are part of the , it's become quite a pain to set up.\nFixed in\nThanks! Note that, to get this to run for me, nixpkgs needed an overlay specifying that all packages wanting would get . That looks liked: It seems that, while is deprecated and python 3 is \"the default\", the package maps to , so a lot of packages effectively default to python 2. Thanks to for coming up with that workaround. I haven't yet see the docker integration work, but I've seen the plugins and such install, so that is promising, and it's likely the last bits are in my config to sort out.\nare you on an older release of nixpkgs? has been set to for the datadog integrations since .\nWe determined the evaluation error came from here: Did we mess that up somehow?\nThere's a good chance that can be removed from there with no impact. We recently did so for the systemd service.\nis a vestige of the old agent. It was a python wrapper that launched that jmx java thing. can most definitely be removed check is broken. I imagine we'd need to package"}
{"_id":"q-en-nixpkgs-c933f4f4627e0be8a70ff0748c7cdb7dfb78daed541b37aa2e7b423abaef9f59","text":"Project description Gut is a CLI designed to make Git easier to use. Metadata homepage URL: source URL: license: mit platforms: linux, darwin, windows"}
{"_id":"q-en-nixpkgs-6d509a7c189c1230511205ee338fde8c6476c18dd83dc14b2ded6fd9f6691e79","text":"libnfc from the cache does not seem to work for me: ~ What happens is a null-pointer dereference somewhere inside IOKit: ~ I am really not sure what is going on, but what troubles me is that I cannot reproduce the build on my laptop: apparently, the examples depend on readline, which is not listed as a dependency. ~ How does the manage to succeed on Hydra? What’s wrong with libnfc or libusb built on Hydra? After I as a dependency the build succeded and the resulting libnfc is fully functional.\nDo you use sandboxing?\nI honestly do not know how it works and if it is enabled by default.\nI have opened to address the readline issue (which is probably completely unrelated).\nI ask because it could account for the difference (that is, there could be an impurity in the build, causing it to break in unsandboxed environments EDIT: or allowing it to succeed, whichever). It's not enabled by default, but I'm not sure about how stuff works on MacOS wrt sandboxing.\nHm, you have a point. I guess it’s not sandboxing but something similar: building in a pure nix-shell yields non-functional utils.\nOk, now I see. Without it found Homebrew (as the expression does not depend on and therefore the system is not shadowed) and it told it about libusb installed from Homebrew. So, apparently libusb is the culprit.\nOn my end, I've found that with to the build environment, the readline error manifests (whereas it builds fine otherwise); adding both & allows it to build. There's no segfault, however. As a side note, FWIW, is much weaker than the isolation provided by the build sandbox (it just clears most of the shell environment, you can still traverse the rootfs & pick up all sorts of impurities).\nHow did you build this? Do you see the same result with ?\nTry running: in your nix-shell. see\nWe need some proper way to fix that.\nWe could have it set in at least temporarily. There are no real side effects to just setting it besides the potential impurities.\nI spent most of yesterday trying to debug the issue but with no luck: I did not manage to build libusb and libnfc with debug symbols (I have no idea how works on macOS; apparently it puts debug symbols into but does not copy them to the resulting binary, therefore, when the library is installed into the store, debug symbols are gone. It would be great if someone taught me how to do this properly).\nWhat exactly is that impure thing that makes pull system CF?\nIn this case it's the framework, that's linked against the system CoreFoundation, I'm still testing some stuff but I'm pretty sure that I have a proper solution."}
{"_id":"q-en-nixpkgs-aaf5ade694e636cd837b527f59da77c1492e789c2f8f69bec07ae1f693b8c49b","text":"Steps to reproduce the behavior: rustdesk None\nShould be fixed in , thanks if you can test it.\ntested and it works! Will close once it gets merged lands on unstable."}
{"_id":"q-en-nixpkgs-9ac4014ccc23024510d32e5641827b3c73bed968f251e68aefad800ad00af8f7","text":"Package name: django-sesame Latest released version: 3.1 - <!-- Search your package here: --- Current version on the unstable channel: 1.7 Current version on the stable/release channel: 1.7 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR.\nhttps://django- The docs for django-sesame ensure compatibility with Django 3.2 and 4.1, which are the versions covered by unstable and the latest stable releases. There is no guarantee that 1.7 would have been compatible with the current Django versions."}
{"_id":"q-en-nixpkgs-a2afb686e29d72bcd89b19eed2e60dcdd10f26413a27acb4e2bdd47567bb53f2","text":"Installing and using fzf breaks with the following error message: Steps to reproduce the behavior: fzf with tmux integration, I did it via home manager modules fzf with tmux integration within tmux The tmux pop up appears with history to filter If applicable, add screenshots to help explain your problem. Add any other context about the problem here. Please run and paste the result. Looks like this got introduced with The runtime dependency was with that new version: Line 183 of file I wasn't able to find a proper URL to link to that line within the diff.\nWorkaround: Install as another package.\ndue to providing the update to 0.39.0 at\nShould be fixed by , would be helpful if you could test it.\nThanks I'm pretty new, is there any doc how I could test this on my machine?\nIf you're using Nix channels then you can replace your channel with and rebuild, which will build all your packages from my PR branch, including . If you're using Nix Flakes then you can change the nixpkgs input to be which has the same effect."}
{"_id":"q-en-nixpkgs-380b0eafee0a7b3afbb3635c39ab3d2d22abcf7a6fe268044ffe96eda47fcfef","text":"It would be nice if the packages provided a icon. the following in your : up a graphical application launcher and search for Wootility. Please run and paste the result. <pre<samp- system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: </samp</pre\nPR submitted that, among other things, adds a desktop-entry file."}
{"_id":"q-en-nixpkgs-0a27b110658a56c47d97f50171f16c4ffe26c12be5193731453d31919ed0ebd2","text":"Codux is a visual IDE for React that enables users to visually edit their projects and see changes reflected instantly in their code editor. It aims to streamline the development process and provide a more intuitive way to design and develop React applications. <!-- Describe the project a little: --Metadata homepage URL: source URL: (unfree) license: unfree platforms: windows, linux, osX"}
{"_id":"q-en-nixpkgs-a347b9bf91bad4a77b59e7694b6e1476149e3f920dcf4edde2636bb59c355248","text":"does not appear to be supported on ; this is possible via , and works with no issue, so it does not appear to be a technical limitation. I am not familiar enough with Nix to port over the homebrew packaging, but was planning on taking a stab if this issue does not get traction. or and System: Nix version: Nixpkgs version:"}
{"_id":"q-en-nixpkgs-35715b7ca258c0187cb9d41d391d10dfb8ff992c6e671369fdb7b45ef4124522","text":"Currently, it asks for password four times to create a hotspot, such as reported at It was fixed in I think in NixOS, polkit is disabled by default, so I enabled it with , but it does not seem to make a difference. Please run and paste the result.\nThank you for the bug report! Which command are you using to create a hotspot?\nI'm using the gui . It asks for password once. Then enter SSID, Password, Wifi interface, Internet Interface, go to Advanced and check 2.4Ghz. Press Create hotspot. It now asks for password thrice.\nAh, again it's linux-wifi-hotspot assuming it's located in /usr/bin"}
{"_id":"q-en-nixpkgs-1fda54e597a9edfca19805dae67c90cd4877df0b6816636e24f2694c29aa3599","text":"When typing something here: <img src=\"https://user-\" width=50%The desktop crashes and you have to re-login (well it's Chinese in the screenshot but I hope you get the idea) <img src=\"https://user-\" width=50%Steps to reproduce the behavior: budgie menu something crash No crash. See above. I don't see this issue before the GNOME 44 update. Probably another issue but I haven't go deep yet. I am on right now, it's nixos-unstable with pull request picked.\nBacktrace points to <details</detailsLooks like there are changes on that touches after 10.7.1, I am testing that right now.\nI gave the latest HEAD (i.e. ) a try and cannot reproduce the issue so probably some of the commits fixes the issue. I wonder if this can be reproduce by others and if some fixes can be backported before 23.05\nOhhh goooooood and now I am reading , let me make some noise there Edit:\nAnd now I am on , which has glib 2.74 ( has 2.76). This is going to be weird because with G_SLICE=always-malloc I also cannot reproduce the crash (budgie-desktop 10.7.1), so probably not related to the slice allocator. Edit: and it's vala 0.56.4 on (no crash) and 0.56.6 on (crash)\nOkay I guess I can basically confirm it's caused by vala update because I am now on and cannnot reproduce the crash, wow it's first time I know valac can really cause issues :seenoevil: Bisect is too painful for my slow laptop :upsidedownface:\nCould you try compiling and running this Vala code on the broken version to see if it also crashes? It should do the same thing as the crashing Budgie code, just without crashing your desktop. If this ends up being a Vala issue, we need a minimal reproduction program to send. Build and run with (assuming you put it in a file called ):\nI am now on and my desktop crashes, so the current range is :upsidedownface: I am away from keyboard for the next few hours because it's midnight here but feel free to drop questions :upsidedownface: I built the program with vala 0.56.5, it exits normally with\nWell, crap. I was hoping that would crash. Maybe this version when you get back? It's closer to the code in Budgie: Of those commits, the only one that really looks suspicious to me is at a glance.\nI am still here (for now) :laughing: but this runs normally for me as well (hmmm can we really guarantee \"if I can trigger the desktop crash, then using same vala same laptop same setup we can trigger a crash with some different program\"?)"}
{"_id":"q-en-nixpkgs-e86785152aa4642fd629febdaa8eb9fe00cbed85e6ab3aea7ea8e8c740709a19","text":"One uses to create a new tree of files for generating a web site. So the intent is to edit those files. But the files produced are read-only, so user must do a before each new file can be edited. I haven't looked at the source yet, but I imagine this is because jekyll copies the site template out of an installed gem or something, and all the files there are 444 (read-only) because it's in the Nix store. System: NixOS 16.09.1600. (Flounder) Nix version: nix-env (Nix) 1.11.6 Nixpkgs version: 16.09.1698.\nHave you tried reporting this upstream?\nI have not. Will consider that – it seems that a fix would be a one-liner after to do or so. But I'm not able to test easily if that's all that would be needed – the nixpkgs expression is quite behind, carrying jekyll 3.0.1 when current release is 3.4.1. I tried updating too, but ran into bundler problems, and I don't know much about ruby.\nYou seem to have idea how to fix it. Should really bring this up with upstream. I don't know much about ruby, but nixpkgs manual has a step by step guide: Turned out to be very easy:\nSo I ran into bundler problems)\nOk, I've pushed a workaround for that"}
{"_id":"q-en-nixpkgs-95ea27a748f5c4e2b3a4861fd6a56acfd97de04a394d6cdf08101b379191e329","text":"Package name: tracker-miners Latest released version: 3.5.1 Current version on the unstable channel: 3.5.0 <p</p- The package update itself is trivial, but I got stuck at addressing . There is now a ioctl check that runs: The program seems to build fine, but does not exit with 0: This makes the Btrfs subvolume feature disabled: I am not sure where to get start to enable this since in they decide this will be a hard check on next major bump (well, it should be possible to opt-out but it should be possible I guess). :seeno_evil: Note for maintainers: Please tag this issue in your PR.\nI am getting , even though I do not use BTRFS so I assume this depends on the kernel version of the builder? (I am using .) That does not really sound like a good idea since users can use completely different kernel version than Hydra. Maybe it should be an and the check performed only when it is . Though I am not sure what the consequences are when incorrectly configured is used. Maybe it will need to be a runtime check instead?\nI saved the program as a separate and compiled & ran with as root user and it exits with 1 for me as well. says I am on 6.3.0. I use btrfs so as root runs fine for me (if that matters). Hmmm then this is weird since it sounds like is introduced (is this the correct place to check such info?). But later I found in the BTRFS subvolumes feature is disabled as well :thinking: Edit: Looks like the fedora link is broken,\nApparently, it falls back to success since I do not use BTRFS: (both inside and outside the sandbox) <details<summaryModified <codetest.c</code</summary</detailsSo that does not really sound like a good test. Edit: I .\nHmmm now I wonder if their original idea is to check whether the code , probably knowing is there is already enough? (I assume they are not supporting < 2.6.34 kernels then I expect all linux should have it?) It looks like they already has some .\nCould you try running the modified in build sandbox to find out why specifically it fails for you?\nThis is using and putting in : This is not in builder sandbox and run as root:\nThe fix is now merged upstream:"}
{"_id":"q-en-nixpkgs-930189d20a98916de4d2ed1e68d62b4a10e2bcf0dc03e680cfa71c4c481eeea2","text":"Given you have PHP installed on your host and one installed in your Nix shell + symfony-cli, eg. with these packages , running will always prefer PHP installed by host. This is not an issue with Nix per se but symfony-cli, but I suggest here that symfony-cli package should be patched so that this does not happen. Steps to reproduce the behavior: Terminal 1 terminal 2 Curl statement should return PHP/8.2.5, as that's the one installed by Nix. Not 8.1.2. Current workaround is replacing symfony-cli with which strips out everything {php}/bin Please run and paste the result.\nThat does not make sense. Why does Symfony not use the correct PHP from PATH? I would expect any proper PATH respecting program to return the same thing as .\nI agree after making a couple of tests, I believe that wrapping the binary with the PHP path is not necessary... Check the screenshot in\nI don't know, but this is what they use for discovery of php versions I've not dig depeer than having a brief looks though. And I have to remind that this is about command. command does look to be honoring $PATH correctly.\nDoes this means that does not employ the same mechanism for locating in the and commands? which seems totally uncommon. It would be nice to have the confirmation from one of the maintainers ( ?)\nI've prepared a patch at I hope they will merge it soon. Otherwise we should consider adding a patch for this here.\nI think this issue can be closed now.\nYup, thanks again!\nAfter thinking twice, the issue will be definitely fixed as soon as you publish a new tag and we provide an update in Nix, I'll leave it open in the meantime.\n7 has been released now.\nPR is made, I'll wait for the CI to be done before merging and closing this issue."}
{"_id":"q-en-nixpkgs-113a40baf8abd2bc4499477e84bb05d9d013e06dcf25c92ffcd18aa8e9b3870e","text":"Package name: Latest released version: <!-- Search your package here: --- Current version on the unstable channel: Current version on the stable/release channel: <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-e89a4b3db112f8d5c2f2be7cf62082630e1c421163002d7670c61e7e79ee7fab","text":"Project description A simple GTK wrapper that installs and updates the skin. Metadata homepage URL: source URL: license: gpl3 platforms: linux"}
{"_id":"q-en-nixpkgs-13ae7f365937d351ffd5412c389ea5180fa0fb21e91b328c7cf4c523e54fb259","text":"The steam client (beta) is attempting to access gsettings schemas and fails. Exporting manually as fixes the issue. into the (likely don't have gtk3/gsettings-desktop-schema already addressable on the host, not steam-run, path) steam the following error GSetting should be looking for schema files in BuildFHSEnv moves schemas to XDGDATADIRS in the fhsenv doesn't include is being pulled into the fhsenv via It's possible just adding to the end of the environment variable is the fix. I'm unfamiliar with gsettings. Attempting does not work as in the fhsenv is missing things adds in . Please run and paste the result.\nExporting GSETTINGSSCHEMADIR manually did not work for me. I left beta participation by removing the file:\n+1 affected by this. Edit: this results when you participate in Steam-beta. This is from recent update - however recent major update worked perfectly (new sleeker interface and all that), breakage appeared around 24 hours ago and it was pulled by Steam. This breaks for all users in system. Removing beta flag (as in McDread solution above) works, causing Steam to downgrade. Upgrading into Steam-Beta after this will cause Steam to break again.\nSetting GSETTINGSSCHEMADIR worked once for me, but no longer after a reboot.\nTry setting it in the extraProfile attribute in steam.override like this\nI was able to fix this by overriding steam, adding and setting :\nThis works great! Tested with reboot.\nThis worked\nFor me, this was in the location . Removing that file worked.\n+1 affected by this. workaround does fix it."}
{"_id":"q-en-nixpkgs-491084485d63eac944e29788cee1136c1c4d044642e6ab7aedd8b40bc709a4c2","text":"I'm using , and I'm trying to perform an update: The update works. (information found from )\nWhat does have to do with ? Can you reproduce the problem with just & ?\nBeginner here. I think that is using under the hood. I don't know how. Should I report this issue to instead?\nSounds like breaks vscodium darwin build :upsidedownface: I will do a PR"}
{"_id":"q-en-nixpkgs-61994e74a2d7239015845caf051a36096bc5ff84c6cb09017e94c5b6b7b9dabf","text":"Project description A command line editor for HCL Metadata homepage URL: source URL: license: mit platforms: darwinarm64, linuxarm64, windows_arm64 Release page: Copied install info from readme: If you are macOS user: Download the latest compiled binaries and put it anywhere in your executable path. If you have Go 1.20+ development environment:"}
{"_id":"q-en-nixpkgs-ff0d746ebad0929b2730ac1b3483c1dea1e0b60d19086b6e26c71bb805fa585e","text":"IntelliJ (and I'm assuming other jetbrains products) uses a Python script to restart itself whenever needed (e.g. a plugin update), and if it detects that there is no Python, it shutdowns instead. Steps to reproduce the behavior: intellij. In the menu, select \"File\" -\"Invalidate Caches...\" there is no python in path, the opened window will have a button \"Invalidate and Exit\" run from , it will have \"Invalidate and Restart\" Same thing happens whenever it needs to restart, with/without python it shows \"Restart\"/\"Shutdown\" buttons respectively, and in some conditions it just asks \"Restart now?\" and then does not come back up when you agree. I know it's easy to fix on my end, but it's not super clear that that's even an issue, and I think python should be to dependencies and made visible to jetbrains IDEs. And yes I don't have system-wide python installed, that's the whole point of NixOS, is it not? :) I guess\nI can confirm this affects the other JetBrains IDEs as well. Adding to the solves this problem. I'll open a PR for it shortly."}
{"_id":"q-en-nixpkgs-d4a8267c002d18325d5d43f72e6cb262827fbb528ab8ef6b99339b84315c39a2","text":"Project description Interact privately with your documents using the power of GPT, 100% privately, no data leaks Metadata homepage source license: asl-2 platforms: all?\ni am really new to nix, and i made this by adding dependency until it worked.. but hopefully a good starting point now just need to do is, which will download the models and to run it."}
{"_id":"q-en-nixpkgs-36c39d823929801232f4daa556f02765771b5c95ddb3f7b4c962c2977c75ccb9","text":"A clear and concise description of what the bug is. You can browse files and it'll show them but playing a song (in my case an MP3 file) results in this error: On the next start, the song is greyed out. I've d it and it seems to try (and fail?) to load libavcodec. That might be why. is in spotify's library path but somehow it's not loading. Investigating now. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\nFallout from , PR incoming."}
{"_id":"q-en-nixpkgs-d48af55b0bb3fb3ad7f1adca278662aa8aa2e1c0137b2427327b07d0c0096e78","text":"When opted in to the latest Steam Beta (version ) starting Steam fails with the error Looking at the log output, it looks like it fails to load : Steps to reproduce the behavior: into Steam Beta steam Steam runs ! stdout when running : Please run and paste the result.\ni'm able to reproduce this as well\nIn case anyone experiences this and wants a temporary workaround, you can opt out of the Steam Beta by running\nI think this might be due to a change in nixpkgs rather than steam, since this works for me on: But it breaks for me when I update to:\nBuilding with running with and using the bad commit is part of cc\nPlease test ."}
{"_id":"q-en-nixpkgs-eb61111f4713b41d1c329e78643829b29634071f82a41c75aefebb94c9cb0f75","text":"Steps to reproduce the behavior: Error: There should be a working matplotlib for 3.8, and the newer one for 3.9+, I guess Commit caused this regression. For 3.7.0: Latest matplotlib that worked for 3.8 is 3.6.2. by simply taking the old version of (before the said commit), which builds but fails to work with tk (I'm not sure tbh how to approach it). Please run and paste the result.\nalso cc\nMaybe, try poetry2nix or using older nixpkgs.\nI tried with but it gave me a error when trying to compile some boost library...\nSorry, but I don't have the energy to provide full support for the python package sets we don't recurse (3.8, 3.9) into. As such they're untested and the fast pace at which python libraries move makes it cumbersome to keep everything in working order.\ngreetings ! lists python 3.8 as compatible ( sorry you need to look at html title tag to see that its for version 3.7.1 matplotlib builds with this change, tested at I appreciate how much work you put into this, so I am happy to create a PR for this you should be able to just override attribute of derivation, you can refer to nixpkgs manual for that, or let us know if you need some assistance\nThank you! you can try now to use the master branch of nixpkgs with the shell you tried before\nThank you very much everybody. This fixes it :)"}
{"_id":"q-en-nixpkgs-a08bdbd47bb90c971aad083aadbac69cc97714e8df7ff4caf20f42eac6a9c6b8","text":"Since , was moved to top level, and and are aliases of from the top level To find the usages of these aliases, you can search for in the repository Note that there are false positives (e.g. when used in overlays) and this also might not include all usages (let inherit)"}
{"_id":"q-en-nixpkgs-4f4e6602c18e366a05e759d305584d6adf267af77f1473e5ac1d2e3017b6d781","text":"Project description This is a CLI keyboard layout switcher for Gnome 3 and 4x. It is not based on the X interface but rather implements direct D-Bus messaging with the Gnome Shell. Metadata homepage URL: source URL: license: bsd platforms: unix, linux, darwin, ..."}
{"_id":"q-en-nixpkgs-c829112a9b931418d9809581da116abc6946a133fc5c8abfdb1441b38202e883","text":"Project description aw-watcher-window-wayland is a part of ActivityWatch needed to have it working in a Wayland compositor. At the present time, without this, the ActivityWatch package is not working on Wayland environments Metadata homepage URL: source URL: license: MPL-2.0 license platforms: linux"}
{"_id":"q-en-nixpkgs-ae567e53ab16abdb06791d46c5c14a4235df5857030409df5d321ee41dc27013","text":"now has an available file (). There are two implications here: should no longer use a generated/patched lockfile can use now because is avoided (which means the custom can be deleted as well) n/a n/a\n, I ran the update script but ran into\nThank you! I'll keep an eye on it as well.\nIs this going to work with still open?\nit would still run, but the package would fail to build, so r-ryantm wouldn't make a successful update until is merged"}
{"_id":"q-en-nixpkgs-f89a0c5d37e47d02ae264ed6d6971007fd9832373d58931ac8de2b5149e56208","text":"is using the memory GSettings backend on my unstable NixOS system. Because of that it forgets any change in its settings done by the user. The window manager used is . The system configuration has in . Before the latest change () was to in an explicitly wrapper, which was replaced by . There is no explicitly mention of anymore in the nix expression after this change. Can it be related? System: (NixOS: , Ubuntu/Fedora: , ...) Nix version: (run ) Nixpkgs version: (run )\nYeah, this is related -- my fault. Thanks!"}
{"_id":"q-en-nixpkgs-b5b52486d38eb54c1eb9cc0a710275830a3cf86ba58c0f75a327d7f329611c90","text":"Project description linux-wallpaperengine aims to reproduce the background functionality of Wallpaper Engine on Linux systems. There already is a wallpaper-engine implementation packaged, but it only works on KDE, while this project should work with most Wayland-compositors and window-managers. Metadata homepage URL: source URL: license: gpl3 platforms: linux I tried to package it myself, but the \"just compare with other packages\" approach of the nixpkgs documentation was very confusing to me. If someone can point me to other documentation or anything helpful I'd be happy to try it again.\nSort of, it still runs under X fyi My rough process: check the repo of whatever you're packaging to see what language and tooling it uses read the nixpkgs manual for the language and tooling that you're building with find examples of other expressions in nixpkgs that use the same tooling to get an idea of the structure use something like to generate the boilerplate for the expression , get an error, use to find the required dependency (repeat this step until there are no more build errors) submit PR"}
{"_id":"q-en-nixpkgs-50464205702d46eb067d2208a086d1c173c307bd90a0003188f09e9e312f0f1d","text":"Applications are starting to allocate more times than allowed by the default value of vm.maxmapcount, . For instance, this situation arises in applications such as Ryujinx () and newer Windows games like Hogwarts Legacy (). SteamOS has already addressed this matter by augmenting the aforementioned value, while Fedora intends to implement the same modification in Fedora 39 (, ).\nPlease feel free to send PRs, it's hard to track those issues as you don't mention any owner.\nThe line to add is: The questions are: How big should n be? ? ? In which file should I add it? At the bottom of\nYou mentioned other distros were thinking about bumping it, let's follow Fedora lead In the relevant file, a quick grep shows indeed that sysctl is one of the best candidate. :)\nFor the policy discussion: I'm all in favor of doing this soon(m). It is not really dangerous and as people have explained to me the previous default has been there for historical reasons and 32 bit systems."}
{"_id":"q-en-nixpkgs-e74717216072e8d825772673ac903e367c9a0125b9a14f8551510ff929cd5662","text":"Originally spotted by (thank you!). Steps to reproduce the behavior (assuming and are both ): The environment variable contains semicolons and must be quoted, otherwise bash will interpret the semicolons as command separators.\n- Hey, I am not sure how to clean up the mess I just made on . Really embarrassed. I intended to rebase and hit merge instead, which pulled in 10k changes.\nNo worries, Happens to everyone at some point :) Personally, I'd just open a new PR.\nHere it is: . Does the commit history look clean enough? I'll try to rebase it later, but I am concerned about retagging all of those folks.\nI'd say don't worry about it -- as far as best practice goes, I'm actually not sure what's recommended for commits. do we recommend squashing before merge? I know I typically just do force pushes to avoid any commit history :l\nI went ahead and squashed, and I'll post a nixpkgs-review when it's finished.\nI don't think we have an official stance on this. Contributors use their best judgement for whatever keeps the commit history simplest and easiest to use"}
{"_id":"q-en-nixpkgs-66b24d323a1846d10c4db66a51ff605311ddaeb57c330db5ffd40849025e5322","text":"the sway test is flaky, both because ocr itself seems to flaky and because ocr seems to be very slow. Steps to reproduce the behavior: a couple times, we haven't gotten it to work even once this seems to be blocking nixos-unstable channel updates. there are a number of timeouts in the hydra build logs that we can very clearly attribute to either ocr not working, or ocr timing out because it gets only one shot at finding the gpg prompt (and if that fails, loops forever on the empty desktop).\nI found it failing quite reliably, exactly since merging the .\nscrolling throug hydra logs it's been flaky before too though, something is definitely up. there's failures in early may that look exactly like what we see now\nWhere do you see this? Looking through I only see successful builds (apart from the cancelled builds and dependency failures). I also don't remember it ever failing when working on it. So this definitely seems to be a regression from staging-next (either it just causes issues with (some of) the OCR tests or it can even affect users - in which case it would be very good that the test is failing). Unfortunately I don't really have time to look into it right now (maybe Thursday or Saturday). It could be due to a font rendering regression. Hm, that's strange - the GnuPG pinentry-gnome3 passphrase prompt shouldn't disappear automatically / after a short timeout (if that's the case than we definitely need to change that part).\nBoth and failed consistently at this point: So it looks like either doesn't start when using the Wayland backend or the font rendering is broken then (with the X11 backend (XWayland) the test succeeds).\nSome of the “cancelled” builds are cancelled due to it timing-out, e.g. and ; the reason there may be unrelated, though. Indeed most appear to be non-timeout cancellations.\nrun locally here the things the tests wants to happen did always happen, but OCR was incredibly flaky and either did not detect them at the first try, or not at all. this is especially bad for pinentry because it closes after 10 seconds for reasons we have not dug into. it looks like this is squarely a failure of OCR (and the way it's used here does seem at least somewhat suspect, maybe using other tooling to list windows would be more appropriate? would almost certainly not take three minutes per attempt to find a window)\nAh, right, thanks for the examples :) Indeed, that seems to be to the gnome3-pinentry timeout that I wasn't aware of (so that part would be fixable). Weird, that shouldn't happen. OCR is slow but should be reliable. Not sure if this could be setup dependant. On Hydra it doesn't seem to be an issue. That is indeed bad. Luckily already found a solution for the timeout in . I intentionally chose OCR to ensure that the (font) rendering works but it isn't required for ensuring that the windows are there. The slowness of OCR is certainly a drawback though."}
{"_id":"q-en-nixpkgs-a528ee04ce3b6a88c3f1b078c1a2edec9a7802b23ed488082a7d7bd6f6ef6088","text":"Project description pyenv lets you easily switch between multiple versions of Python. It's simple, unobtrusive, and follows the UNIX tradition of single-purpose tools that do one thing well. It is a critical application for Python development. Metadata homepage URL: source URL: license: MIT platforms: ALL"}
{"_id":"q-en-nixpkgs-9174dc35858b99ced037db86e4c9287281e6382f9f6ddfd459f7abc3be7a378d","text":"Spideroak GUI does not work in NixOS unstable. When running , I get this error: As far as I am able to tell, headless mode runs fine. Install spideroak and try to run . System: 17.09pre102667. (Hummingbird) Nix version: nix-env (Nix) 1.11.7 Nixpkgs version: \"17.09pre102667.\"\nPython uses the default zlib which is currently 1.2.11. I see the following options: override Python expression and use an older version of zlib let spideroak use our version of libpng. cc\nWell the package's approach is strange to me. It installs private versions of libs that depend on libs not installed (e.g. freetype), and at the same time, some of the missing libs naturally depend on libs that are also installed privately (on zlib, for example). You can try such tweaks to find what will work now, but I'm afraid it's likely to remain fragile."}
{"_id":"q-en-nixpkgs-2859ee80ae87944f1712c2a367e6eac6322264a83e2214ab424717821ec34899","text":"Project description A fork of Golden Dictionary which fixes many old bugs, provides improvement, and is being actively developed. It might be better to switch to altogether instead of continuing to maintain . Metadata homepage URL: source URL: license: gplv3 platforms: linux\ndoesn't seem too inactive now: they even had a new release last month (based on Qt WebKit, though).\nI think it might be a good idea to have both versions as provides extra features and bug fixes if that's possible."}
{"_id":"q-en-nixpkgs-f0663144546388c77b7fecd240a02cdbb0de63784923690d5efe014f416e75f8","text":"reaper can't export video (wrong ffmpeg version) in nixos 23.05 Steps to reproduce the behavior: the package reaper -Render : \"Video (ffmpeg/libav encoder)\" is missing There should be an option in the list where you can use ffmpeg as encoder If you override the ffmpeg dependency with: (reaper.override { ffmpeg = pkgs.ffmpeg_4-full; }) then it works."}
{"_id":"q-en-nixpkgs-2e0330a21da66c4d2ae2dd6b895821f325988992c978efdb51b63126d93a6666","text":"Package name: ultrastardx Latest released version: Current version on the unstable channel: 2021-04-03 Current version on the stable/release channel: 2021-04-03 Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-5efa933b301863076a55ea70ddf89947306445c21187e265e4e3c7d0cb590c61","text":"Steps to reproduce the behavior: davinci-resolve Uh, also while checking, in the source link for davinci-resolve points to instead of Please run and paste the result.\nI tried building off of and it seemed to build fine How are you building it so I can try to replicate the issue?\nI updated my home-manager flake, and the reported commit hash it now uses says its from 2 days ago so I just have to wait then. Thank you.\nAccording to the PR tracker (), the PR behind is available in But when I try to use it using: ( is currently at ) I'm getting the same error, do I need something else to try it?\nI just ran your nix build command without error. What version of nix are you using? Also can you put more info about the curl error you're seeing? Just want to make sure its the derivation and not nix being the one throwing the error. I figured I would try too but that seems fine also\nI was also on Nix 2.15.1, testing with 2.16.1 give the same error.\nWhen you have a moment can you post the full build log -\nAdding an echo of the resolved url immediately shows the issue: And the log is: Maybe the registration payload changed?\nRunning into the same issue and error message myself\nYeah, is right. Even the where a lot of this is source from is suffering from the same error. Trying to see if I can get it to work rn, stay tuned\nAh, they a street requirement for the form haha. Im going to set it to the chamber of commerce in Utrecht since thats where the Nix foundation is registered\nShould be able to build off of:\nNow I can DL the sources, but uncompressing fills up the build folder and ends in a error :/ (which is seemingly not catched as the execution seems to follow afterward) I tried setting to a folder on a partition with lot's of space without success :/ Do you know how to avoid this? <details<summarydf -h</summary</details\nHonestly im not sure. Checking my it looks like the source is about 3.0GiB So it filling up your at makes sense.. Reading up on some other issues I think this might be what you are looking for if you are on NixOS, just adjust the tmpfs size larger than the default 50%:\nThanks, I confirm that works, I can build davinci-resolve (now I just need to find a suitable GPU.... )\nI tested 's patch to the package, and it wasnt working for me. I found out that the Host for the second curl request to blackmagic's download site was mismatched, so I updated the host to their new site and also updated the street address to the actual location of the nix non-profit (found on the KVK's business lookup). I've tested on all (3) my machines and it builds successfully so if anyone else here would like to try, that would be great. Heres the PR: And my fork is at #davinci-resolve\nJust tried rebuilding today and having the same issue. Looks like blackmagic moved back to the host and have blocked Maybe we should just remove the host header entirely from the curl command? If we are having hashed outputs verified, then it wouldnt matter which host it comes from?\nYeah, sounds like they may be doing some sort of canary release for new infrastructure or something. I agree, that makes sense to remove it.\nShould I submit a new pull?\nSure! Ping me and Ill give it a review. Thanks!\nNew PR at\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-cef011f0e386b535819a6ece7a3cb5c98193b67b1fc6947bf8c9c5f47dbe757f","text":"Package name: MuseScore Latest released version: 4.1 Current version on the unstable channel: 4.0.2 Current version on the stable/release channel: 4.0.2 Notify maintainers Note for maintainers: Please tag this issue in your PR.\nThanks for the notification! Your review is welcome at ."}
{"_id":"q-en-nixpkgs-290791c4824339ae1c0fd98dc7087e73a266c6b11c9c26b492433ca424779877","text":"Project description Language Server for Perl Metadata homepage URL: source URL: license: artisticlicense2_0 platforms: unix, linux, darwin, ..."}
{"_id":"q-en-nixpkgs-80a29b81ac553c9afc2380c19384ac51de62edf326acdd88d6d3c6b33672c72d","text":"I've been trying to see why vscode had rather laggy scrolling and high cpu usage when scrolling on wayland with set to 1, and upon discovering and , I saw that the library isn't found by it, causing the app to fall back to software rendering. Disabling the env variable and starting vscode like that shows that on xwayland it does start up with gpu acceleration. Doing the same thing described , VScode does start with GPU acceleration working. and no longer complains about the library Steps to reproduce the behavior: VScode with set and with the flag, so it starts with wayland support and outputs details about the startup and other stuff how it complains about the library. to see check that it's indeed using software rendering. The library is found, there are no complaints, and GPU acceleration is enabled. I am on an Intel laptop. Please run and paste the result.\nWhats the exact log? vscode vendors that library and it is located in our package at\nRelevant lines from [(44)] Failed to load GLES library: cannot open shared object file: No such file or directory [(186)] Exiting GPU process due to errors during initialization [(44)] Failed to load GLES library: cannot open shared object file: No such file or directory [(186)] Exiting GPU process due to errors during initialization\nSame error happens on VSCodium where running with on Wayland frequently crashes on startup. ing yields a lot of lines before the logged error, including this particularly interesting one: The file doesn't actually exist, only does, so apparently VSCodium is looking for while only is vendored. If is to the , is found as a symlink to the actual library and ing again confirms that this library is loaded by VSCodium. Edit: Adding to fails with different error messages, so probably there's something broken with the vendored library:\nI would guess that's because the drivers to not abi match. I would recommend to rm the vendored shared object and symlink libglvnd to the correct location. Also we should investigate if we can remove/replace the other vendored libraries.\nAdding to the library path did the trick for me, thanks!\nI can't get it to work with vscode sadly. vscode = (super: { libPath = lib.makeLibraryPath [pkgs.libglvnd]; });\nHow soon will the changes get into nixos-unstable?\n2 days max ?\nYou can follow the PR here:\nThis is still broken for me: In fact, that file is missing. is missing too. As expected, due to last PR merged. I can confirm that the wrapper script contains: So what gives? How did this work for others but not me?\nI'm guessing the last PR fixed it for wayland/ but broke it for others? Edit: I can confirm that I have gpu acceleration when I pin vscode nixpkgs to which is the parent commit of . I also tried the following, linking libglvnd's libGLESv2 into vscode directory. The error changed and terminal rendering completely broke with a black screen:\nplease open a bug !"}
{"_id":"q-en-nixpkgs-dd96eb428a5f6ca89f9d0c33808afbe8949829e9fe10e88500fdf286c48770eb","text":"Self hosting server for Rust Desk remote desktop the rustdesk client app is available, but it would be great to be able to self-host. Metadata homepage URL: source URL: license: GNU AFFERO GENERAL PUBLIC LICENSE platforms: linux, windows, mac, android, ios\nWow, that was speedy. Thanks!"}
{"_id":"q-en-nixpkgs-7a5bff41f47e3df62c4fbff36e2fbaffefc32fb7934895088f1a9b0ecfaa117b","text":"Project description An unofficial GoXLR App replacement for Linux, Windows and MacOS. Metadata homepage URL: source URL: license: MIT platforms: linux\nMoving conversation back here to avoid spamming goxlr-utility maintainers :sweat_smile: I made a draft PR that you can find from It's closer to what the official .deb and .rpm packaging is doing, i.e. including some of the relevant resources, shell completion etc. I wouldn't mind adding you as a maintainer but to do that it probably requires adding you to the maintainer list as well. Should be easy to fix if you want to add PR to the branch.\nHey wow you are fast :D Yes, you can add me as a maintainer for this package. But as I understand it, I need to add me to the maintainer list first, right?\nHah, had some free time during the night and since I wanted to get my GoXLR setup working anyways thought I might as well sort this out. For the question: Yup, the requires to be on . The file itself lists the mandatory/optional fields in the metadata, but basically \"a name and at least one contact/identity information\" (e.g. a github id). Oh, and the list needs to stay in alphabetical order ^^ Feel free to make PR against this branch, I'll merge it as soon as I notice the PR :)\nAdded the module to the PR as well. I guess it might be hard to wait for someone with GoXLR to come in and review the PR (though I'd prefer someone to check that it doesn't violate any major guidelines) so if you don't mind I think it's possible for you to try it out and do a PR review, which likely speeds up the process at least :)\nManaged to get the PR reviewed and should now be code-wise idiomatic enough to pass. I think based on the main thing remaining is getting thumbs up from you functionality-wise so that's it's not only \"works on my machine\" :sweat_smile:\nI saw your PR and tried using nixpkgs-review to download it to my local machine and try to run it. I used these steps: Do you know if my approach is wrong? This is the first time I've tried a Nix PR. Unfortunately, I haven't found much information about it. EDIT: Got it working\nGlad to hear you got it working! :) The nixpkgs-review tool seems nice, hadn't used it myself but for building the PR it seems to do exactly what one would expect. I guess in this case there's a limit what it can do though as the udev rules are being on the module side and it probably doesn't handle that so some manual work is required :shrug: But again, all is good since you did manage to test the build!"}
{"_id":"q-en-nixpkgs-1927c61dbeb899b753ad3e67348a6898b761a4faecfcb97e296f76ba3a8a10e9","text":"Another Problem:\npatch welcome !\nin , the call to generate completion should be moved to before the wrapping. And it should be called for luarocks-admin too thanks for noticing. Let me know if you wanna do it.\nIf u wanna test"}
{"_id":"q-en-nixpkgs-1bf7a10e413049839d5c83f3189d2a5eee9aa3dd7c6ffc019123ef0605ebf420","text":"A CLI program that renders Markdown timeline as HTML Announced: <!-- Describe the project a little: --Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin, ...\nSince upstream has a this should be pretty simple using ."}
{"_id":"q-en-nixpkgs-ce3360b11fe9980097367264de1908dc4837d895470fd38631033d1a52c6750a","text":"bunx is missing in the bun package. - command must be present. - Please run and paste the result."}
{"_id":"q-en-nixpkgs-9749fce4e3cd607af09373e6de71562f80341e9d3ef07d5d9d6e081d468f13a7","text":"Package name: expressvpn Latest released version: 3.52.0.2 Current version on the unstable channel: 3.25.0.13 Current version on the stable/release channel: 3.25.0.13 Notify maintainers Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-6ab153dd4546375e3495620636f28bebbf96cffff3353c8689f20d56d8bc9bbf","text":"I feel like the documentation isn't helping a lot on how I can use , or any other alternative library/packaging solution. I've been trying to figure out how to use - the project I'm trying to nix-ify is . issue should give an insight on what problem I'm dealing with - it's about the package . Alternatively, I've also tried replacing with , but then, it complains about the package missing. The three packages that I'm in need of: - -[x] checked [latest Nixpkgs manual] ([source][nixpkgs-source]) and [latest NixOS manual] ([source][nixos-source]) [x] checked [open documentation issues] for possible duplicates [x] checked [open documentation pull requests] for possible solutions [latest Nixpkgs manual]: [latest NixOS manual]: [nixpkgs-source]: [nixos-source]: [open documentation issues]: [open documentation pull requests]:\nHi Ashvith10 - lisp-modules doesn't use Quicklisp or ASDF's - instead, all systems are in the Nix store. You can discover what systems are packaged in : Then, you can build a wrapper like this: Then, you can load the systems like this: Notice that ASDF is 'ed directly instead of using - this is to ensure the same version of ASDF is used on run time as was used on build time. seems to already exist - I don't know if that's the version that you need?. and are not packaged yet. I'm not familiar with Lem build system but looks like we have to package a bunch of extensions For examples of how to package, see . You can start with , , and and work your way through the errors.\nthanks for answer. However, looks like I'm in the wrong here - I should have updated my progress and the misinformation in the top post. and are the packages we're trying to make. The first one is the text editor for terminal, and the second one offers an interface using SDL. The issue right now is when I try to run the build script, it fails because QL does not exist: As you can see here: This is the file we're trying to run to build: which is trying to run this file: Since we didn't need this (because the very packages were updated to the same commit in nixpkgs as mentioned here in the patch), I got rid of it in my personal copy of the project, them through the Nix packages. Now, here's where the problem starts: and Since ql:quickload isn't working, it fails here. These \"packages\" are in the root of the repo, and I'm not able to figure out how to add them.\nWe could try to avoid this build script, and use asdf:load-system instead. There were two other places with references to QL, I patched them out. To add Lem, we can do something like this : For now I'm running it like this: There is some warning about missing font, I skip it with space. I need some help packaging the extensions, it's a lot of work\nthanks for helping out with the project, if you could just let me know what I'll have to work on, I'd love to add my own commits. But I'm still not able to grasp why we're creating extensions and adding them in ? and should only be exposed to the lem package, right? Wouldn't it be better to create asdf build system under the file? About the packages, here's what I had in mind - there will be three packages to choose from, lem (which will include both lem-ncurses and lem-sdl), lem-ncurses and lem-sdl. I'd rather that opt-in be provided for users. What do you think?\nlet's not discuss more about here, it's my bad that I didn't create a separate issue. I'll create one shortly under \"package request\" and tag you there. This was with regards to improving the documentation, especially in cases like this, where instructions from isn't clear, and we might want to improve on this.\nWith regards to the first few improvement I can think of is: In the above snippet, the symlink is missing. and Both of them don't work, if I just type it on my console. I was however, able to figure out how to make the last one work - to and then run the command. There might be other such incomplete parts in the documentation that I've forgotten, because I've not touched in two weeks.\nthanks for catching that, I fixed those issues in commit regarding , I can see some further issues with it: It says that packages are 'imported', but doesn't say to where - perhaps we should say up front that the result of the import process is the file (this is mentioned further down, but maybe should be said earlier). Maybe also say that the goal of the import process is to save people some work of writing Nix expressions by auto generating those expressions and hoping they just work - I don't remember the numbers now, but a majority of them do work. It mentions the untracked \"\" file, but doesn't say that the file can be deleted at will It says that the maintainer should \"re-run\" the script, but doesn't explicitly say when - it's mentioned that the script pulls the latest data from Quicklisp, but maybe we should explicitly say that the script should run when there is a new Quicklisp release. This seems to be an issue with the section. In an ideal world, we just import the Quicklisp data, add native dependencies and everything works, but some packages decide to do weird stuff like creating files other than FASL during compilation. Other packages simply don't exist in Quicklisp. For these packages, we override them in Wchichever you prefer - though the convention so far has been to package things as one ASDF system per expression in . Then apps such as StumpWM or Nyxt (or Lem) can use things from as inputs. The advantage of having it in is that it will appear in all supported CL implementations: , , etc. If you place it somewhere else, you don't have access to , only to , which is specific to one CL."}
{"_id":"q-en-nixpkgs-d3ea649ffb24296d38351f3110d7ad3ccfe73aedc2876949803f9eb050927681","text":"Project description Volume control applet for Pipewire, written in Rust. <!-- Describe the project a little: --Metadata homepage URL: source URL: license: mit, bsd, gpl2+ , ... platforms: unix, linux, darwin, ..."}
{"_id":"q-en-nixpkgs-67e493da612255c1a81e3841ecd7efb8a4b47509668c91f7d623f9166ff668f7","text":"Package name: cava Latest released version: 0.9.1 <!-- Search your package here: --- Current version on the unstable channel: 0.9.0 Current version on the stable/release channel: 0.8.3 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-6dacc4fc23edfec4d93bb05e102581797a242f286f7a6da6460542cbad5ffbb1","text":"NixOS on boot loops waiting for devices with wrong names. According to device paths are split by \":\", but for some persistent block device names we have \":\" in valid paths. The boot is stuck trying to load: Steps to reproduce the behavior: NixOS using /dev/disk/by-path devices boot waiting loop Boot without waiting for nonexistant devices ! Please run and paste the result.\nThis is a recent phenomenon I've experienced a few times, on bcachefs. I'm not entirely sure why this happens, but since Linux 6.4.8 I believe, I haven't had this issue. There are two reasons why this could happen AFAIK: is messed up between the linux kernel and stage-1. needs to have all the devices loaded before trying to mount multiple. That's the best I have ATM. Current nix-info for reference: system: host os: multi-user?: sandbox: version: channels(mado): channels(root): nixpkgs:\nI should also clarify in my case that it was a breaking bug, where I would have to reboot because one of the devices never loaded and would drop me into an error, until it worked.\nit seems like you're describing a different issue with bcachefs and multiple devices separated by . This bug, as far as I understand, is about non-bcachefs block device paths that contain incidentally (not as a multi-device delimiter). I'd suggest filing a separate issue for your problem.\n(But as the author of the change that introduced that behavior - - I suspect you might want to be involved here regardless :P)\nMy bad, the usual bad reading comprehension and I didn't look at the paths that were listed closely enough. I would suggest, if possible, using something like instead until this is fixed. As for the fix, while it would break the use-case of colon-separated if it is the aforementioned path.\nThis should be considered a bug. There are a lot of paths in that have colons in their name (including RAID UUID devices). Requiring users to know that devices with colons in their name are treated special is unreasonable, especially if it's not documented. If bcachefs requires this then I think we need a new configuration option or the stage 1 function needs to be extended so it considers the file system type before setting .\nIMO, we should get rid of it for bcachefs (it's only use) and just use the new ` mechanism.\nI don't use bcachefs so I don't know what the mechanism is. Is it safe to simply revert your patch to ?\nUnfortunately no, because many people's auto-generated configs still use the ':' separator. It's something I think we'll have to do at some point though. I think the better option may be to omit a warning if it's being used (i.e: using bcachefs) until 24.05, backport that to 23.11, and remove it in 24.05.\nI would suggest that we change the function to take an optional second argument, the file system type. If that argument is set and it is set to \"\" then (and only then) do we set . Of course, this only works if the auto-generated configs correctly set the file system type. Can you point me to the commit where you changed the auto-generation process?\nShould be fixed in"}
{"_id":"q-en-nixpkgs-5021ade412b0335dccd3873d71a2a1e3166ec8f33bb5e3e3ec3a78e600f1b35f","text":"Project description zcfan is a Zero-configuration fan control daemon for ThinkPads. Metadata homepage URL: source URL: license: mit platforms: linux"}
{"_id":"q-en-nixpkgs-dcf99485eedab63b6e5f2e051f5564f3bec5c7a94f0b2347e31fba07d8022419","text":"Project description Streaming media player for Plex Metadata homepage URL: source URL: license: proprietary platforms: linux Unfortunately, this app is only available as a Snap or a Flatpak. It looks like Snap is easier to unpack, as we have . There is also , which appears to be unmaintened.\nwould also appreciate this being packaged, the flatpak version does not start for me ( I don't like using flatpak anyways).\nIt also supports Darwin and Windows.\nJust to be clear this is not this package This is a completely different package with different build steps that just looks the exact same.\nSubmitted a draft PR that adds this to nixpkgs. What's there now works on my machine&trade; (3 of them). I'd appreciate people testing this out on different hardware and suggesting any improvements that makes this more reliable. I have a couple more things I'd like to do before officially submitting this. Edit: Also need help with Darwin. I'm guessing it doesn't work there and I don't have a convenient way to test that platform."}
{"_id":"q-en-nixpkgs-309639a76dee32a15fc0ca41d333483902a29fba897fba61c3074743cb5c3e9a","text":"Project description onedriver is a native Linux filesystem for Microsoft OneDrive. It contains GUI and CLI tools for mounting Onedrive cloud shares as local drives, downloading files accessed on demand. Metadata homepage URL: source URL: license: gpl3 platforms: linux\nI am very excited about this one, I am actually a DM in a D&D campaign and onedriver has always been my best tool for rapid storage and edit of the stuff. Thanks everyone!"}
{"_id":"q-en-nixpkgs-5c6a7c2dcbcf256a9b2441fa69b0522962872dc8245456282414dec713783f4c","text":"Steps to reproduce the behavior: itch system: host os: multi-user?: sandbox: version: channels(root): channels(username): nixpkgs:\nLooks like there is an issue already posted for this package NixOS/nixpkgs and there was a PR (NixOS/nixpkgs which uses a canary build instead) posted. Since there is no stable update (the canary builds uses a separate config directory, which is also explained here ) this is still a upstream issue."}
{"_id":"q-en-nixpkgs-bdf2cc640d958d80cc377192bd38b41d3a114cedd63e3d93b2f7cb25dc1d9344","text":"After a recent Nixpkgs git update, my Firefox browser can no longer download files to the local disk. At the point where the file browser is supposed to pop up to let me choose the directory in which to save the file, the entire browser just crashes. The console doesn't say much: This issue occurs in and earlier. I'll try to find a commit that causes the issue by way of git bisect, but that's no not easy to pull off (for obvious reasons).\nThere are and ; there are and . Adding some combination of these to could help. I guess the wrapper needs updating…\n~This one should be just IIRC.~ I was wrong.\nI don't understand why it doesn't crash on 17.03.\nThe fix worked! Thank you very much."}
{"_id":"q-en-nixpkgs-c848066523700b849e80e010407980b70d5d15fedbbc08b5e2dbc9592666f5c3","text":"Steps to reproduce the behavior: that build exited with success code, but directory is empty. Removing the following fixes the issue for me. Disclaimer: I'm not familiar with golang so I'm not sure what this does or why it was in the first place. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result.\nthank you for your bugreport, we need cgo because a dependency needs cgo"}
{"_id":"q-en-nixpkgs-da13e0e4e3d27e5f241b08087a0b8f26e89a4db5e3f82b7f904be7fdd2f607e8","text":"A clear and concise description of what the bug is. Steps to reproduce the behavior: Try to configure to work on all keyboards. Note that, although this is a common use case, trying to do so with the exposed API leads to errors. For instance: This code throws the following It's expected to have some way of declaratively configuring kanata, such that a given configuration block works for all keyboards. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nDoes setting work for you?\nYes, that seems to work! Thank you :)"}
{"_id":"q-en-nixpkgs-75925e50ebbfb7235790efab366c9986b63a6e672896e651b807bba72c3361b0","text":"Project description Floorp is a firefox fork. Metadata homepage URL: source URL: license: MPLv2.0 platforms: windows, macos, linux"}
{"_id":"q-en-nixpkgs-7cd280e6aee7bc900697f3c43bec6466524e74b3168f2232cd586fe6e2f9cd6b","text":"Project description A wayland native, highly customizable runner. <!-- Describe the project a little: --Metadata homepage URL: source URL: license: gpl3 platforms: linux\nNB: they have a flake\nYou have to compile it if you add the flake and the Cachix currently doesn't work. I rather it to nixpkgs because it way more dependent here.\nI can work on getting it into nixpkgs shortly.\nOo nice :3 Thank you so much"}
{"_id":"q-en-nixpkgs-cca8f2b95e8d008dab98d2d9f6bca7078f13b03404a31900b5cab7fe60ee72c8","text":"Project description A Steam Achievements Manager for Linux Metadata homepage URL: source URL: license: gpl platforms: inux"}
{"_id":"q-en-nixpkgs-83a27a31bb42fd76952988c39be102828bd68da5ef827c9963e84554709fd642","text":"When the tooling is updated, it should be run on Nixpkgs in the same PR. This also relates to This could be achieved using on the derivation which copies Nixpkgs to the store and then runs the tooling on it in a derivation. This would then automatically be run when a commit with the prefix is in the PR. This is however fairly slow and expensive due to the Nixpkgs copy.\nOfBorg performs checkouts all the time. I don't think this is significant. To run it locally, this style of test runner is not great, but we can refer readers to the instructions for a manual run.\nIt's not just a local checkout though, if we use derivations to run the check, then Nixpkgs also needs to be copied into the store. Should probably be fine though, other parts of Nixpkgs also copy Nixpkgs into the store.\nI don't think this should be done after all. With the latest idea , the latest version of the tooling in Nixpkgs needs to stay independent of the pinned version used for CI.\nAgree that for the PR checking, the currently implemented approach is fine. However, I think it is still good to have the \"simple\" version that doesn't refer to other Nixpkgs / channels as a attribute. The point of this is not to catch CIs, but to just sanity check with hydra after the fact that everything worked out the way it should. The impure tricky business with channels is still necessary, but the idea is that it is an optimization and the simple self-contained version is the reference implementation that ultimately defines the behavior we want.\nImplemented in , turns out this would've been really useful :sweat_smile:"}
{"_id":"q-en-nixpkgs-c57fd5431de97fabdd57098d45b62608d4ac545b725a6d865793ee23401d02c0","text":"Steps to reproduce the behavior: / / All plugins show the same error with a different hash when building, I only included the log from : The es were updated with but the plugins fail with a hash mismatch on two machines I tried it on. Both are x86_64 machines running NixOS unstable. I can make a PR to update the hashes, but I want to make sure I'm not missing something. Please run and paste the result.\nI've just opened a PR to fix this, if you have time to review, it would be great, otherwise no worries. I won't wait too long before merging since everything is broken.\nLooks good to me, those are the same hashes I got."}
{"_id":"q-en-nixpkgs-db308fd5214a3daf83864bb3a091c38ca6ab5490e6a8bdedde0a3c67315dadcd","text":"Package name: sudo-rs Latest released version: 0.2.1 <!-- Search your package here: --- Current version on the unstable channel: 0.2.0 Current version on the stable/release channel: N/A <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-4e0e60fc6b48fdc7efe3452ae2eb63110d32eaad77cffe9d2006a5a1141f0a09","text":"When creating nixos-containers with umask 077, files get created with the wrong permissions, and services such as dbus don't work because they probably can't read certain files: Like for nixos activation, the creation of nixos containers should force umask 022."}
{"_id":"q-en-nixpkgs-ae738f4a2aa5c43505f8b9e9af140318a6dc0febcf8585bcfcf89cd78030e97f","text":"A clear and concise description of what the bug is. launches, but cannot select language in the welcome screen. If I open the \"preference\" from tray, no text is displayed. Steps to reproduce the behavior: Ctrl+Shift+I to open devtools, found error. A clear and concise description of what you expected to happen. The app should display normally. If applicable, add screenshots to help explain your problem. ! ! Add any other context about the problem here. Running on NixOS unstable. The log from devtool: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\nThis is still an issue, and i think the current maintainer is\nI have tested out the newest version and found out that the above mentioned issue does not happen on this version."}
{"_id":"q-en-nixpkgs-a6a4c52e595ce41effed6b063ea39b7ecc45750614901ee1456ee5ee85415761","text":"Project description A self-hosted BitTorrent indexer, DHT crawler, content classifier and torrent search engine with web UI, GraphQL API and Servarr stack integration. The DHT crawler is bitmagnet’s killer feature that (currently) makes it unique. … DHT’s lesser known feature is that it allows you to crawl the info hashes it knows about. This is how bitmagnet’s DHT crawler works works - it crawls the DHT network, requesting metadata about each info hash it discovers. It then further enriches this metadata by attempting to classify it and associate it with known pieces of content, such as movies and TV shows. It then allows you to search everything it has indexed. This means that bitmagnet is not reliant on any external trackers or torrent indexers. It’s a self-contained, self-hosted torrent indexer, connected via DHT to a global network of peers and constantly discovering new content. Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin"}
{"_id":"q-en-nixpkgs-cbfa0a8ffcb688ba98797584ec2776b87df1762d4300cb909f0d1bafe0ddd196","text":"I note that the package sets in .[(1)] says [(2)] that the reason to disable \"JBIG\" is patent concerns, that disabling it removes support for many printers, and that \"The patent will expire on April 4, 2012\". Given the last, need we still disable this feature? [(1)]: [(2)]:\nI don't think so. openSUSE enabled it, too. Other distributions might have done the same, not sure.\nA PR that implements this change request would be very welcome. I'd rather not have this issue lying around, though, without someone having committed to actually work on it."}
{"_id":"q-en-nixpkgs-b50f8606e8d03440f7b32cdd9496a209786bfd59cb737e6d8b8f14dbe19c1d00","text":"The home-assistant component currently declares that it's missing the input However the package actually exists, it's just called . This appears to be the old name of the package, it looks as though it got renamed at some point, although the nixpkgs packge for it wasn't updated for the name change.\nNote that merely setting restores the component's functionality."}
{"_id":"q-en-nixpkgs-d340a380fc95575e8b2895044d2b919de6e0c11c56f21e4554356dea77463c07","text":"After babashka can't display Unicode characters I have a babashka script that outputs the following JSON: The text is a Nerd Font Symbol code point. After the output is: Steps to reproduce the behavior: This has been fixed in the past by I did not test other packages that could be affected by the PR.\nPretty sure this is related to this bug in GraalVM upstream: I am curious if this is affecting projects using GraalVM 21.0.0 vanilla or if not, how is it working and why it is broken in .\nIt does seem that even setting up inside the C compiler environment is not sufficient to make it work. CC to see if they have more context on what the possibly issue is.\nI also hit this today. I tested the same version(v1.3.186) from upstream binary release, which does not have this issue. Shall we always sync the same version of graavm as upstream build ci?\nI don't think this is the issue and even if it was, it is not like babashka is the only GraalVM package we have.\nI know that folks can't reproduce this issue with upstream, but I also can't reproduce it with GraalVM using Java, e.g.: A patch if anyone wants to test to be applied to : So, can anyone reproduce this issue with GraalVM Native Image itself? If not, the only thing I can think is that the issue here is the GCC (i.e.: non-static) version (upstream compiles with by default). Or something else.\nSo I got this warning: But this was after already started to build, so I assume this is the same issue as and need to be fixed upstream.\nTo explain in more details why I think this is an upstream issue: I don't get the same in the initial part of the build, only when the build starts that I get this warning. So what I assume is doing: Running the CC (C Compiler) inside a shell with clean environment We see the , because this shell couldn't set the proper locale thanks to missing environment variables (e.g.: ) Since this shell environment has no environment variables (at least, not the environment variables that we need), this ends up breaking macOS compilation and locale (Unicode) support in Linux Not sure why it works in GraalVM itself though yet. This is all really strange.\nWhile I'm not familiar with graalvm, I do find a hack for this issue, maybe can weigh in to test whether this hack is valid: Some inspiration from a random reddit post and those issues: And I do find it's using , so I thought if I those two into build args, maybe it'd work. And indeed it seems to work.. After I built bb with above args,\nBTW, I cannot reproduce with GraalVM Native Image on helloworld java as well. Helloworld produced by GraalVM Native Image has correct unicode output.\nThanks for investigating this issue. Opened so we can have this fixed."}
{"_id":"q-en-nixpkgs-5c5c2ae34f22e24584ca51d6659a55542ec00a552a5196db991af1eb4029af7a","text":"Project description Python REPL plugin for automatic time tracking and metrics generated from your programming activity. Metadata homepage URL: source URL: license: gpl3 platforms: unix, linux, darwin, ..."}
{"_id":"q-en-nixpkgs-06afa5a8eacaf0eb1cddf135259581001c7885563b5dd9111a570755a7938cec","text":"Package name: python311Packgaes.openrazer Latest released version: 3.6.1 <!-- Search your package here: --- Current version on the unstable channel: 3.5.1 Current version on the stable/release channel: 3.5.1 Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Last update pull request was made by: Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-a7ac24ce52c08e6fcc690bf9e633489633f685c01e9329e3bb7f981390de0e53","text":"Looking at my filed ofborg does not ping in most cases. I am thinking of ofborg pings people, or the above command does not print (I guess) No response. No response."}
{"_id":"q-en-nixpkgs-da827a178373c2309c3b3e2ac4b45a1d1861bee150a0c75a00d3d54c9d63a50c","text":"Package name: rubberband Latest released version: 3.3.0 Current version on the unstable channel: 3.1.0 Current version on the stable/release channel: 3.1.0 Notify maintainers Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-a96e353f40b65250ad348c20ba2276242b1e6193de5c110c2ab8ca0097855303","text":"Project description A powerful testing framework for ZSH projects Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin, ..."}
{"_id":"q-en-nixpkgs-000b5e0422e8df6969c79251d9cde1d90b2f0d3151ad0c02602dfaaccfb6e58a","text":"rendered in : ! bumping osl to did not work. Applying did not work <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --- system: host os: multi-user?: sandbox: version: nixpkgs:\nI have verified that the problem exists with the current derivation even with version , which was updated away from in . I guess somehow the ZLIB_VERSION is replaced improperly , but I've got no idea what could be causing it, and I'm really confused why it's only become a problem recently, as it doesn't look like osl has been touched in either.\nI've tracked it down to , and reverting it seems to fix it. I also believe it's a cmake issue. Consider this file Output: Maybe because the header changed to (? But if I'm reading the correctly, I think it should just set the version to an empty string? Either way, this patch seems to fix it I don't think it's worth raising with zlib because with the next patch version they're going to fix it anyway. It might be worth warning them in advance of 1.4. It's probably worth raising with cmake, because they will want to support all version, and then zlib won't have to care anymore either. I can open a PR for this tomorrow or when I have time, but don't wait for me if anybody wants to do it before.\nNice catch! One minor note: you are looking at 's master branch. It seems to be already fixed there as: We'll need that backport to . UPDATE: or wait for to reach master. With now is in ."}
{"_id":"q-en-nixpkgs-c2d04f803692d38e20574189c2ff21d4fa99472b8290ef33634a6979feace2e7","text":"Package name: raysession Latest released version: 0.14.2 Current version on the unstable channel: 0.13.1 Current version on the stable/release channel: 0.13.1 Notify maintainers Note for maintainers: Please tag this issue in your PR.\nUpdate PR:"}
{"_id":"q-en-nixpkgs-de427abb95997039513b3f6bdd80662b727e1a908ad04540769acde183bbbea3","text":"A NAT port forwarding rule fails to build with nftables when ranges are used. This is also the entirety of nix log /nix/store/sir8pvbk65zk4j6rqcj8sgwi0y78s3kc-nftables- Steps to reproduce the behavior: The nfables-rules derivation should compile. Using a single port rule, or disabling nftables results in a successful build. Please run and paste the result."}
{"_id":"q-en-nixpkgs-1f2793925c15ccab0d9405c688301e432cae1be8e4b914911368318d39f68e50","text":"H.264 videos and specifically those (WebM and Theora are fine) don't play, which means that videos on Facebook and Twitter (at least) report an error. There are instructions for getting a patched version of the ffmpeg lib installed into your home folder which works, but ideally wouldn't be needed: to that the H.264 video don't work but the others do. System: (NixOS: , Ubuntu/Fedora: , ...) Nix version: (run ) Nixpkgs version: (run ) Sandboxing enabled: (run )\nWhat browser are you using? Works on Chromium, Firefox and qutebrowser (qtwebkit) on master for me.\nVivaldi, hence the title of the issue...\nOh, sorry, didn't know there's a browser called vivaldi."}
{"_id":"q-en-nixpkgs-de4ed279fed0328641857308b5fef4a5519d004ef7735a2b0fc7096fb974e770","text":"Steps to reproduce the behavior: macOS: Since , this fails on macOS due to case-insensitivity (). The offending PR had no description, no reviews and no reports of tests on any platform. It was merged by the author four days after it was opened. Please run and paste the result.\nFuck MacOS and their stupid case insensitive file system.\nCan you try mv'ing the file first to another name like?"}
{"_id":"q-en-nixpkgs-977dbb13870de8c2a6511fc5766fb52bdf71b6056a34b486502f2790fde5d03a","text":"Project description A sleek notification daemon for Wayland compositors. Metadata homepage URL: source URL: license: mpl platforms: unix, linux, darwin, ...\nPR:"}
{"_id":"q-en-nixpkgs-1eefefc2746de04a976091cca86705044076b00622400c105758b6614f8f2a91","text":"Steps to reproduce the behavior: Posting on behalf of someone who was asking for help on Discord. Please run and paste the result.\nBad hash was introduced in\nLooking at the ofborg history on the PR, I'm wondering if upstream did a sneaky tag change shortly after the release...\nsneaky tag-switch theory pretty much confirmed. Can we switch this package to always use commit ids in the future? If they've done it once...\nbefore switching to the more obscure commit ids, I would ask upstream not to do this anymore\nI think it's better to notify in upstream so he's aware of the situation.\nI already notified about this in so he's aware Meanwhile I'll change the commit hash to match the new id.\nMy bad. I was caused by the GitHub Action, so I force-pushed the tag . Really sorry for the inconvenience and I will never do this again"}
{"_id":"q-en-nixpkgs-2d624f37aab674deeb4f86a61722a78ab5cfe5e894e15b8978df73c6fcc0ce8c","text":"torchWithRocm hydra build fails because of too many logs. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --\nHow exactly do you fix this? Is there a hydra option we can put in the derivation? Otherwise, do we have to omit warnings?\nI have a fix by omitting some warnings, but should we do that? Should we instead get hydra to allow bigger logs?"}
{"_id":"q-en-nixpkgs-4eb9a73b9ec7e7d2cedcbbdd73306196af4f37574fb2e91bb2668405a4fbb397","text":"Package is broken at version 0.1.8. It still worked at version 0.1.7, so I assume that is to blame. I've verified that the issue is fixed in 0.1.9, so I assume all that's needed would be to kick the bot. Not sure how to do that. No panic"}
{"_id":"q-en-nixpkgs-38e38bb0a50b005bfa7554cbea050dd7da8f19077bb88453d1cbe4cc381f68d9","text":"Steps to reproduce the behavior: vulkan-validation-layers Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\nThe dependency causing this is vulkan-utility-libraries: 1.3.269 -1.3.270 This version bump causes vulkan-validation-layers to fail. Its only dependency is and that is fine. So problem is\nThe latest SDK version is 1.3.268. vulkan-utility-libraries should be rolled back to that version; we should not be packaging non-SDK versions of Vulkan SDK components, and especially not piecemeal.\nindeed reverting it to fixed the build and runtime"}
{"_id":"q-en-nixpkgs-ef7795ba9b43ee43cd1028b921261d31c5d41f8316052910eaf011243a86971e","text":"root has not been successfully built since 2023-08-27 with version 6.26.10 by hydra."}
{"_id":"q-en-nixpkgs-a84673e39329ed1ce202f1a5a27c462331c28fad6bfeb3c9ee87978425af5dd7","text":"When specifying multiple paths in a single parityFile entry (SnapRAID ), the systemd service failes with the following error: Steps to reproduce the behavior: a SnapRAID config and specify a entry like so: run the snapraid sync: the error: The runs without error. SnapRAID has a where you can specify \"split\" parity files. This is useful when you're using 16tb or bigger ext4-formatted disks for parity. ext4 doesn't support files bigger than 16tb so this \"split parity file\" can be used to specify two parity files on a single raid disk and SnapRAID will automatically use the second file when the first cannot grow anymore (hits 16TB). You specify these multiple files by putting multiple comma-separated paths in the . This mostly already works except when it comes to the scheduled systemd sync job where it specifies . If you specify a parity with multiple files you'll get an error when the systemd job runs: It seems like since SnapRAID essentially does not allow commas in the parity file paths it would be OK to split on comma here? Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-6e15d2a2f302722a81a9aef998f6ea4763cac6a9986c1b1b338fe277a6520a27","text":"Budgie desktop stopped working a few months ago on because gsettings schema supplied by Budgie is missing settings which was recently to . This causes to crash during startup making Budgie DE setup unusable. budgie-desktop from . .\nAre you using Budgie alongside another desktop? Otherwise LightDM is the preferred choice for Budgie, but I will still look into this.\nOh, I didn't know that. When I switched from gnome to budgie, I kept and everything worked, so I kept it. I can confirm that if I use , everything works."}
{"_id":"q-en-nixpkgs-704acc3bdad4e97ec87656e5b08bbe2d2cd03c49cf3fc70783d0a532f8ec6554","text":"After the upgrade to 0.68.0 fsautocomplete no longer attach properly to NixVim (neovim). Steps to reproduce the behavior: a minimal NixVim config enable fsautocomplete and Ionide-vim a .fsx file the LspInfo command a see the missing id The clients get associated with an id as before. Version 0.67.0 ! Version 0.68.0 ! I haven't explored if the problem persists outside NixVim, but I would strongly suspect it does, as there haven't been any major changes in NixVim lately. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI can confirm you findings. It's an issue with 0.68.0. I'll do some more investigation before filing an issue upstream. for now, you can pin you fsautocomplete to 0.67.0\nI've just been testing it a bit as well and came to the same conclusion that it is an upstream problem. I reconfigured my nixvim to build the from the nuget package and that one has the same issue. How do I pin in nixvim?\nsorry I am not familiar with nixvim, my neovim just calls fsautocomplete from PATH. So I've just installed an old version via thats the commit just before 0.68.0 update\nare you able to test ?\nWorking on it."}
{"_id":"q-en-nixpkgs-10a5b9a2b92f638998a411fce97175a41a5f8ef2a29ba7b4c5edd0c4fe33f300","text":"Project description <!-- Describe the project a little: --I've recently watched the videos on norg and found the plugin very code. At some point though I had an error because my system was missing this treesitter grammar: It doesn't seem packaged yet (I checked unstable). Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin, ... Priorities Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-e8ccbee82832559107e318c7f343c596d1bff7c784e8cbfcce221feae37f9083","text":"Package name: ticktick Latest released version: 2.0.0 Current version on the unstable channel: 1.0.80 Notify maintainers Please update to the latest version. Thanks"}
{"_id":"q-en-nixpkgs-3da26f1d9e4e5f3c8f0168f35b9f12b423be788f7388a8df747ce899625f9b8c","text":"Steps to reproduce the behavior: spotdl I just tried to install it, doing it from or gives the same result. Let me know if I can provide more useful information."}
{"_id":"q-en-nixpkgs-1550046ad86062a1c5609f2da1ec551376955cfce78d8af175cf1c3d32649d91","text":"Upon login, Fcitx5 presents a popup notification telling me to unset and . This is specifically in the Plasma 6 (alpha) Wayland session. The environment variables appear to be set in the Fcitx5 NixOS module. also for ibus. I'm not sure if the other IMEs are supposed to work on Wayland at all. (nixos-unstable with Plasma 6 alpha from nix-community/kde2nix)\nIt's mostly impossible to detect if the user is currently using the plasma wayland session (unless with some custom profile scripts). Maybe we should add an option allowing disabling the automatic configuration of these environment variables.\nFWIW Plasma 6 defaults the \"plasma\" session to wayland. The X11 session is renamed to \"plasmax11\". At least Plasma has some startup script generated by nix, I think at that point it might be known whether it is Wayland or not?\nBut users can still choose the session to use upon login right? However the current fcitx5 nixos module unconditionally sets these variables, without a runtime detection mechanism.\nIs that possible to detect that at runtime, with something like ?\nWell that's the part. But I'd rather let the users decide than trying to act smart. There are both pros and cons of using the immodules or not.\nThat's why I pointed out the startup script from the plasma module.\nI mean it's not a must (or even should, IMO) that you unset and for the wayland session. The input method story is very complicated, with different applications/compositors/input methods supporting different protocols of varying compatibility and functionality. Setting these variables by default in most cases, won't worsen the situation. You may read https://fcitx- for some background.\nHmm, I see. Maybe they could be set with so users or other modules can easily override them?\nReasonable, but empty string and unset do have different meaning regarding these variables. So I think a blanket option to unset then entirely is more desirable.\nIs there any temporary workaround? I can't find out how to unset it, I don't think there is way besides making it blank which might work? I am confused as to how to get IME to work on Hyprland or Wayland at all though, if I can't get Japanese IME to work it's a complete deal breaker for me personally Edit, this worked for me personally. Using Fcitx on GTK crashes my Hyprland, but seems the QT one works fine? I am super confused now. But glad to see that IME support seems to be baked into Wayland itself somewhat I just hope it makes it work just as well as Windows in the far future: Also you will need this at the top of your file where you set it\nIf hyprland crashes, that's a hyprland issue, the fcitx5 gtk immodule works fine on most other platforms, such as sway."}
{"_id":"q-en-nixpkgs-55d2218bca00cd7c17745476aab000ca66680be44df7b7b1c5790155ee2c3c35","text":"when I install spyder with and then launch it, it throws an error with the following traceback+log: This seems to be a manifestation of this issue: where QDarkStyle needed to be updated for the reporter's IDE. Steps to reproduce the behavior: spyder IDE starts As mentioned, this seems to be a manifestation of Please run and paste the result. (Note, I'm using flakes for my os configuration) Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI experience the same. If according to Spyder 5.5.0 depends on QDarkstyle 3.2.0, do we users have to roll back to NixOS 23.05, or could QDarkstyle be updated from 3.1 to 3.2 on NixOS 23.11?\nCould QDarkStyle 3.2.3 be backported to NixOS 23.11 to have a working Spyder IDE? In the meantime, I figured out that I can start Spyder using a . Spyder then gives the message . But I can run in a terminal, use the hamburger menu of the iPython console within Spyder (lower right pane), click on 'Connect to an existing kernel', choose the file generated with , and finally have a working Spyder. However, in my opinion Spyder and all declaratively installed libraries should work out of the box on a stable NixOS release.\nI'm not quite sure how the backporting process works but I can look into it. Personally I got my setup working with this (using ): and this not guaranteed to be minimal"}
{"_id":"q-en-nixpkgs-befcfd0333123e33e644e535aadd412694b86d06bb61aca88311b41957e3bbc9","text":"When building the system configuration, there is a check for the generated nix config, which seems to use internally. This check fails when using with a drv from nix master, as there creates a hard error and exits with a non zero exit code. The message suggests to use instead Steps to reproduce the behavior: to a nix from master. Seems to happen since 2023-11-28 on my CI the system it fail, similar to The build of the system suceeds Please see the linked build logs instead nixos/nix seems to be the PR that introduced the behaviour on the side. (only person touching relevant module) (author of the nix PR) This is the last known to build version: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nUh, I though deprecated aliases were only throwing a warning, not a hard error. is that really the behaviour we want?\nOh, but it's not an error from the Nix side, it's the NixOS module code that rewrites warnings into hard errors :grimacing:\nYeah I have my fingerprints all over the module... The reason it escalates warnings to errors is Maybe we could make the activation script print warnings instead, so we don't have to do this kind of thing. Anyway, the simplest fix is to use the new syntax. could you try ?\nWould the new syntax also work for users that still use nix 2.3? I will test the PR later today nevertheless.\nOkay, just checked the changes in the PR. You do indeed a version based switch of the actual command to use :+1: I will give it a try in a couple of hours, need to be back at PC first.\nYeah, I guessed that would be something like that. I suppose it could make sense to have a way to have these triggered as hard errors (although that would be a bit tricky to implement because that would mean having a setting that influences how the settings parser works :upsidedown_face: )\nI was finally able to check, and the PR seems to work for me. Just took ages as I had to build qemu from scratch…"}
{"_id":"q-en-nixpkgs-3fb92bff4346effdda9c02d901a2897355aabe9a6262995e2b27a54622a4121b","text":"Steps to reproduce the behavior: build certmgr-selfsigned, eg Caused by this by by this out of date shar and also detected Please run and paste the result. Add a to . Thanks!"}
{"_id":"q-en-nixpkgs-aa92fed896a7634da17b448f0873d884dba9141e9d16ac9617498ea504f82f37","text":"The out-of-the-box configuration of Mosquitto will accept any password. The setting is irrelevant. I don't think these details make much difference, but I'll document them here just the same. up Mosquitto with TLS. up a user with some ACLs and a password. to connect as this user. Note that the password argument is irrelevant. The problem seems to be that the option is not specified. An effective work-around is: It seems this is expected behavior. See this matrix at: System: NixOS 17.03 (deployed with NixOps)\nI don't think this fix is within the spirit of the Mosquitto project. The Mosquitto project makes it easy to get started with the service, but you are now throwing that away too. The way I would like to see it work is to have a boolean option \"insecure\" where true means the behavior as it was before the PR. Additionally, when the password file is not set, a helpful hint is given pointing at the insecure option with a big warning about the security implications as well as the option for setting the password file to a real secure configuration.\nSorry, I don't often get time to work on this. I'll try to leave enough information here that you can make decisions and push forward without much more iterations. I wrote some tests: And here's a default-ish NixOS config for mosquitto: And here's the result: Invocation with allowAnonymous --------------------------------- No auth PASS Unregistered user PASS Registered user, no password PASS Registered user, bad password PASS Registered user, good password PASS This is so far from my expectations that I'm inclined to treat it as a bug. Why should the NixOS configs allow the user to specify the contents of the password file when those contents will be completely ignored? Here's what it looks like when we modify the config so that the password file is actually used: Invocation with allowAnonymous --------------------------------- No auth PASS Unregistered user FAIL Registered user, no password FAIL Registered user, bad password FAIL Registered user, good password PASS This is more like what I would have expected. I believe this is not too much of an extra burden on setup. If ease of use is your primary concern, you can either turn on and have your client not specify a username and password, or just hard-code every password to . I understand opinions may differ, and my arguments could prove unconvincing in the end. pointed out this is a breaking change, which is a very good argument against it. So here's my plan. I'll update the PR to include a new option and leave it off by default. If the value is false, you get the classic behavior as seen in the first table above. If the value is true, you get the behavior from the second table. I'd hope that one day could default to true, but just having it in the first place seems like an improvement.\njust fyi, as of right now the default config is secure as no client is allowed to write or read any topic. This is due to a change in mosquitto server where an empty ACL file means no access for anyone. I stumbled over this issue after my upgrade to 19.03. For future generations, this is a setup for a server:\nOuch this last comment really saved my day -- and saved a deep dive into more detailed logs. What I like most about Nixos is how its configs are short, to the point, and sane/secure by default making a working setup. I think we should update the service definition just a bit to make it workable by default. I just can't think up the best way here -- so that it worked for new configs and didn't make any unsafe surprises for the existing old configs.\ngiven the comment of do you think this issue can be closed?\nThanks for fixing this bug :) In my opinion the the description of the checkPasswords option is misleading. It should better be something like „Allow user authentication“. When it is not set to true, even clients using the correct password are refused (assuming the default allowAnonymous=false). I would plead for setting the default of checkPasswords to true, since this is what a new user would expect. If we don't want this to be a breaking change, it could wait till the next nixos release. A compromise would be to set checkPasswords=true whenever allowAnonymous==false. I see no usecase for both being false, it just yields an unusable setup.\nthe next release will have a rewritten mosquitto module () that defaults to checking passwords and not allowing anonymous logins. additionally allows omitting password checks once merged, but will still default to checking passwords.\nFor anyone interested, this is how a configuration for Mosquitto looks like on 21.11 with the rewritten mosquitto module: thanks to on Edit: closing this issue for now, as a lot has changed since 2017"}
{"_id":"q-en-nixpkgs-83fcccd0772d521cad88767b04c2a5567f0f37e1abefd0ce2f06ccc869950c31","text":"Sunshine fails to run when is unavailable () Run , observe warning about failure to load , then a segfault. Adding in an override with fixes it, but I don't know the area well enough to know if this is the correct fix, I just saw this was the package the previous versions had a normal dynamic link against, while 0.21 appears to only access it via . Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks! I totally missed this warning. Working for me too -- opened to resolve."}
{"_id":"q-en-nixpkgs-594a78efe3f6952918dd709a16217c436f01221506873d08acc17077207c73e5","text":"Since Librewolf was updated to 120.0-1 in , the remote seems to have stopped working correctly. Steps to reproduce the behavior: The last command should hang for a few seconds, printing the following terminal output (new as of 120): And then the following message is displayed: A new window should be opened after executing the second command. ! The dbus service still seems to appear, so my best guess is that the problem is on either side of the equation. I'm still trying to work out what exactly is going wrong. If anyone has an idea of how I might debug this further, or has a hunch for what's wrong, please let me know. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThis is an upstream issue: It looks like the dbus name patch was missing some stuff/broken for 120. This should be fixed in 120.0-2."}
{"_id":"q-en-nixpkgs-ab2b469b53102b4cff1193147d5f48fa769f73d48a984e2e6221ef65588c045e","text":"Steps to reproduce the behavior: maubot with sqlite db in nixos config my config:\nthis isn't a build failure (as build failure is for packages failing to build, not module issues), but thanks for the notice. The problem is a missing when checking for local Postresql databases with credentials. It can be fixed by switching to in I'll only be able to open a PR tomorrow, you can open it now if you want to try contributing, or you can wait for me to open it\nsorry for wrong issue type, i'll wait for PR ;)\ni disabled assertions just to test it out and got error during build"}
{"_id":"q-en-nixpkgs-28b3959a92304d162231195ed895d43ed0cb128513b56aee224e26e4eed60c8a","text":"Today provides an outdated set of : Specifically I keep hitting on any attempts to rebuild 32-bit packages. Currently I am patching the to provide the working tools built locally: I think it would be generally useful to be able to pass non-default tools for workaround purposes and testing purposes of new tools. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\npeople who want this, can you review"}
{"_id":"q-en-nixpkgs-4bc3d899d3de3197429424d51c5f6f2c3855a64d8b2c1ffecfc924c3428e7b43","text":"Steps to reproduce the behavior: bitwarden Please run and paste the result.\nWill be fixed by"}
{"_id":"q-en-nixpkgs-e9378f3f9b2f86ff9853fe553b724d6323ad42d6dbc8cf7de84de2813fb8ddf3","text":"at calls Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. at Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-437c4ed8db1222ae032b0f7811855e9a10512106285c56c502976f814799f7e2","text":"use instead of pulling the windows zip from github releases\nI can review if you want to PR this\nupstream has said that they'd rather us use the github mirror that they're going to set up\nthe github mirror has been created\nSorry, I have to delete the github mirror, due to bandwidth limits. Please use\nthe \"SRB22.2\" branch and/or \"SRB2release_2.2.13\" tag\nGetting 500 errors when trying to access\ndoes this work for you?\nThe link works however the download buttons don't. PRs welcome if anyone can get it working, but wait for to be merged first to avoid merge conflicts.\ni have had another look, and have found that it might be better to pull directly from branch\nThanks, was able to get it to work with :+1:"}
{"_id":"q-en-nixpkgs-e15bf2f4b340b69e1e955c2abf080be10cd735d6d289702ed1181c8841339816","text":"Steps to reproduce the behavior: with the following flake: I think this broke with the recent cudaPackages changes, but I haven't figured out the responsible commit. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nNot directly relevant but we should rename -for consistency at some point."}
{"_id":"q-en-nixpkgs-bdc899bebec4988739f8fa591ca1b8f45bd4d58150da8239425612db6e450390","text":"Package name: gnuradio Latest released version: 3.10.8.0 <!-- Search your package here: --- Current version on the unstable channel: 3.10.7.0 Current version on the stable/release channel:3.10.7.0 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nTnx. Update is proposed at"}
{"_id":"q-en-nixpkgs-e174568dff56f21e95d54951e83107cc9aa4fea1fab5a980d42b515dd27d29ac","text":"Package name: gplates Latest released version: 2.4 Current version on the unstable channel: 2.3 Current version on the stable/release channel: 2.3 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-4d38bc18276013a64c6bf9b7807610bc15c455721fc17564345e4b0cf63c0b52","text":"Steps to reproduce the behavior: to create a Julia environment with The issue appears to originate from a missing hash within a dependency. However, the root cause of this problem remains uncertain to me. Building a Julia environment with other packages like and does work flawlessly. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks for the report! I'll take a look."}
{"_id":"q-en-nixpkgs-4b6e7c7293e2376189938d342057150d858cbf33bcc0214908bd2b5abc2537f1","text":"Project description is an approximate model counter for CNF formulas. Metadata homepage URL: source URL: license: mit platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-a125bebc13260856fdaa510aa15fe6769838d0872f0d7bae16059e849fd97fef","text":"Steps to reproduce the behavior: odin If you run the resulting binary through you'll see this strange backtrace: I think it has to do with the version of the C++ standard library that is getting linked. Adding and to the and updating the file to explicitly link allows it to run without segfaulting. (But thats just a workaround I'm using). EDIT: During more testing with the built binary, it also doesn't use the correct linker, which can cause problems if you are linking with Objective-C code/frameworks (not sure if this is a problem with how Odin is packaged with nix, or if its a problem that the Odin devs need to solve). Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]: Note: Was fixed by\nmacOS Odin support is kinda scuffed anyway with any deviation from what's expected. do you have the code for your workaround somewhere? then i can see if i can patch it in\nI'm using for my project. It doesn't have any switching based on OS, and my override probably isn't necessary either. Honestly, the is probably the only thing needed to stop the seg faulting.\nupdating to llvm17 is not a bad idea either"}
{"_id":"q-en-nixpkgs-8ef7241af0c79de7cc17bb5f9937721665a52a4458408a563d6a90e0c5223cce","text":"a canon pixma mp210 nixos-23.11 installed (with gnome) cnijfilter280 to print The print filter crash I do not know how I can debug this.\nhow did you install the package, and how are you printing?\nI installed the package with I tried to print using LibreOffice writer on flatpak. If i pin cnijfilter280 from nixos-23.05, it is working:\nHello, the glibc in 23.11 has more hardening checks to detect buggy code, and this crash looks like the code in the filter is tripping one of those checks. Can you try replacing by and retry ?\nThis is working\ncan you test ?\nThis is working too\ndoes it work with (note rather than )\nYes"}
{"_id":"q-en-nixpkgs-3a7e00fe6864eef88b5e994f11e7eb8b0ecb86977352dcd6035ddaaf82f67520","text":"<!-- describe your problem --The logo at the top of the README is currently using a now deprecated method for using different logo's depending on the user's theme From: <!-- propose a solution (optional) --This is fixed by switching to the recommended method. I believe it should be something like <!-- make sure this issue is not redundant or obsolete --- [x] checked [latest Nixpkgs manual] ([source][nixpkgs-source]) and [latest NixOS manual] ([source][nixos-source]) [x] checked [open documentation issues] for possible duplicates [x] checked [open documentation pull requests] for possible solutions [latest Nixpkgs manual]: [latest NixOS manual]: [nixpkgs-source]: [nixos-source]: [open documentation issues]: [open documentation pull requests]: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-6be30c173ac979de99d9b3fa89569100e7301211eea2202654d31761d63f5ca6","text":"Offline help is not available on Libreoffice, even though flag is enabled by default. Steps to reproduce the behavior: libreoffice to the system packages the offline help is missing the source for a potential setting, notice it's enabled by default and nothing ever touches it. for 3 months unsure if this is PEBKAC or not Have a toggle to build with help, or have a separate package that has the offline help. If applicable, add screenshots to help explain your problem. ! <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- (note: raskin leads to an unrelated org, rename?) cc Please run and paste the result. PS: appears i got a double-zero issue id :D Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks for the report, I think this should fix it (haven't tested the full build yet): .\nI waited for CI to be green in to make sure it doesn't break the build. Could you please test it on your machine, and see if it indeed helped?\nseems like it didn't use the cache so i'm waiting 2 hours for libreoffice to compile... fun!\nHaha yes I know :) That's why I waited for CI to be green - to make sure that at least you won't encounter a build error after hours of building. Ideally, CI would have pushed to an online cache server such as . See also:\nit works. it fucking works. !\nGreat, glad I could help :). Follow the PR at ."}
{"_id":"q-en-nixpkgs-2f811db8759e4b83c46b77e48aba74401c4be9bc5df549e4d5587472065c1f06","text":"Project description Universal dependency update tool that fits into your workflows. Metadata homepage URL: source URL: license: agpl3+ platforms: unix, linux, darwin, ... note that this software may use an open-core model, with features they mention to advertise the paid subscription being 'automatic scaling and Merge Confidence workflows'. edit: an existing nix effort to package this ."}
{"_id":"q-en-nixpkgs-1d63d33cb5ecf336786482386a5f66c0acb225dd7c5588da9d49c9e76b238bed","text":"Package name: nixops_unstable Latest released version: <!-- Search your package here: --- Current version on the unstable channel: 2023-10-26 Current version on the stable/release channel: 2023-10-26 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-8e7f28d08d78a0aa38d06ef3d01ca9762642001dc39725a1aa51ca68613052cf","text":"Package name: visidata Latest released version: 3.0 Current version on the unstable channel: 2.11.1 Current version on the stable/release channel: 2.11.1Release post: Changelog: The new version and its tests seems to want pyconll and a fork of savReaderWriter as new dependencies. The package is also missing a attribute. Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-515b8f39c9174855a5daf65a34094f3e4838c67910d3d8068ae6efa3fa6b1b1f","text":"Agda 2.6.4.1 a configure flag that controls whether debug printing (with ) is enabled (previously it was always enabled). Should we enable this? I am leaning strongly towards \"yes\" as I use debug printing quite often and it would suck to have to rebuild Agda. cc\nis there any downside to enabling it?\nPerformance is , though I'd say not in a way that matters.\nIf it was true by default before, we should keep that default. We can add an option to the derivation to make it configurable.\nOpened I am not sure how to add an option to the derivation (I think should do) and I don't have enough motivation to add an option to the wrapper."}
{"_id":"q-en-nixpkgs-a81c6085cc4a66dbcfe7e00ef842721bf9ee49369a3511ec7fdbc91c0198ad76","text":"Package name: brlcad Latest released version: 7.38.0 Current version on the unstable channel: 7.34.2 Current version on the stable/release channel: 7.34.2 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-2b78612af797bed39d35339052d39d532f9c342b966f93eb91d448ec0ab25053","text":"Package name: unrar Latest released version: 7.00 beta 3 Current version on the unstable channel: 6.2.12 Current version on the stable/release channel: 6.2.12 I receive more and more RAR files which are created with Unrar 7.00 beta versions / require a more recent version tan 6.2.12 I would appreciated if the existing unrar package can be updated or a new unrar-beta package be proposed .. Changelog : Notify maintainers\nThere was an automatic update in that I marked as draft, exactly because it's in beta. I'd much rather add a new package. what do you think?\nI don't think we should update to the beta either. Sometimes I think about adding beta and release candidate versions to package , but that would be a nixpkgs policy discussion.\nIf people are still creating new RAR files then I don't want to be maintainer of this package. Stop using RAR."}
{"_id":"q-en-nixpkgs-dd1499708a2eb146350e20370df8a3efa73c1135aa2c0a68670f7b719a66a4bc","text":"Package name: crunchy-cli Latest released version: 3.2.2 <!-- Search your package here: --- Current version on the unstable channel: 3.0.0-dev.10 Current version on the stable/release channel: 3.0.0-dev.10 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR."}
{"_id":"q-en-nixpkgs-33843686445a041b4c42a6972e78bb23cfe8744bd60d175d503181c7e98699ca","text":"Right now, only supports passing a schedule of dates with or without a . This means that either: machine that uses the same module will update simultaneously () need to override the on every machine to something different (machine a is , machine b is , ...), which is manual effort can build jitter into it with a value of greater than zero, but since every run re-rolls the randomness, updates can now be extremely non-uniform - for example, a system set to update daily with up to 23 hours of randomness may update twice within two hours, and then not for 46 hours. Of course, all of (1)-(3) are adequate enough for most general circumstances. However, systemd 247 introduced which generates a stable random offset (up to ) which is different for each machine ID. This allows for the consistent periodicity of (2) automatically instead of manually. Please consider whether (default ) would be an appropriate new option to add. Until then, I'll be setting it myself through ."}
{"_id":"q-en-nixpkgs-54f4744e42e29f812824861a232d596375e53464b720514f8115e236f1bf1ea2","text":"Package name: limesurvey Latest released version: 6.4.1+ Current version on the unstable channel: 6.1.2+ Current version on the stable/release channel: 6.1.2+ Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-4c76edfe94dc2a2b8bcfef331ff50a0d7e64d22a1aa373ba5651bb84dd84631a","text":"uses two actions and do get necessary authorization via . but these actions are not properly registed, leading to error: (It can be verified that these actions don't exist by using ) This breaks and . I tried both pcsclite 1.9.5 and 2.0.1 ending up with the same error. Steps to reproduce the behavior: and in , configure your smartcard (yubikey) to use yubikey error There actions are registered correctly and usable. Please run and paste the result. I'm using flakes (no channels). Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\ncorrect description of your effect, i am not able to use my yubikey with gpg.\nThis can be solved by\nthis works, thanks! but this is really unintuitive, given that is already adding the package to : I am leaving this open as I believe a proper fix should be provided in the module itself.\nSame issue here. It was working approx. last Dec. I don't have a specific commit hash. Adding to works for me.\nLooks like a regression in () - is implying to me that polkit actions get included in the output, however: isn't included, which explains why adding to fixes the issue. Solution ideas: Add to . Not sure about this one? seems to also contain some libraries that probably don't need to be installed. Install polkit actions to instead of . IMO this one makes the most sense, especially as it seems to have been the previous behaviour before this change.\nI was bitten by this when attempting to update over the weekend as well. In my case it prevented login because I use smartcards for user login which made debugging difficult due to the number of reboots required. I tend to agree with that going back to having the polkit actions file in the output instead of makes the most sense. That will mean that of the pcsc module will again result in the polkit action file being to the system configuration.\nCc: ()\nI also think the location should be fine, the files are also there.\nI ran into this issue when I couldn't access my yubikey anymore via pcsc for commits. The above solution worked for me. I'm going to add some comments for SEO so this is easier to find\nI think you have found the right culprit that caused the issue, but I suppose the issue has more to do with the itself, where it only includes if . This has been rectified in from what it seems. For now the proper solution seems to enabling polkit with .\nI think that makes more sense based on the existing file structure but maybe we just switch bin to out, so it is the default and move the libs to lib instead of being in out.\nI've created to make the outputs a lot more expected by moving the udev file and the systemd service to the out output and the libraries to the lib output: I didn't test it yet, so it is still in draft. If someone could verify that it doesn't break security keys again, that would be great and then please undraft it, otherwise I do it in the evening.\nI do not think I can fully test this due to the amount of rebuilds, so please chime in for reviews and find the one mistake I might have done (which I hope I didn't)\nTried to build but it fails on my machine:\nI've xdist to dulwich here . Let me see if we can just disable some tests as a speed up from 66s to 9s is quite good.\nThis should be a lot better with . I did a lot of rebuilds locally on a low and high core machine and didn't encounter any failures anymore.\nThis was fixed in , works fine for me."}
{"_id":"q-en-nixpkgs-bc6b31538f5370a8196cce201c09684d82c9a3691a7d515c9fd8991385409b67","text":"Project description ssh-tpm-agent is a ssh-agent compatible agent that allows keys to be created by the Trusted Platform Module (TPM) for authentication towards ssh servers. Metadata homepage URL: source URL: license: MIT platforms: unix (at least) Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nah, thanks i promise i did search but somehow didn't get any relevant results. probably PEBKAC."}
{"_id":"q-en-nixpkgs-041d74e0c6d0dfecb3108793e45abfa4fda55be2e32ae8fac27da73804c3c997","text":"Steps to reproduce the behavior: for n/a <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --built from nixpkgs master () Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-7325a08d149b96764836228ef2eac34111065f825cc97a7a3e2d73fb903be061","text":"Ffmpeg 6.1 supports vulkan video, but it is not enabled in the default version of ffmpeg, it is only enabled in ffmpeg-full. The default version of ffmpeg should support vulkan video so that applications like mpv can use it. Arch enables vulkan by . Vulkan video already works with amd GPUs using ffmpeg-full with this command: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThis seems reasonable. I'm in favor.\nEDIT: Error on my end, skip reading this. Hm, your command doesn't work for me on Mesa 23.3.3: Another aspect to consider is closure size.\nIf you haven't tried already, can you try with something in h264? Vulkan video currently only supports h264, h265 and av1(experminental), and only if you gpu supports them.\nIndeed it's h.265, let me try h.264. This looks like a generic VK issue though.\nAs suspected, this happens on h.264 too.\nDo you have multiple GPUs in your system except the amd one perhaps? This chooses the first one arbitrarily.\nSame error for all.\nThen I am not sure about what's causing it, that's the exact same error I receive when using ffmpeg 6.1 instead of ffmpeg-full 6.1.\nUhhh about that... So and that commit didn't seem to have been pulled into this brain thingy of mine... It works as expected with ffmpeg-full."}
{"_id":"q-en-nixpkgs-6cd273a342c85a5de75191ba0288c1f1f5e7c6579a2cfd8098f3d44f4bf045f8","text":"The clang stdenv (e.g., ) on recent unstable channel no longer seems able to find the C++ headers when used as the stdenv for a build. yields Curiously the libcxxStdenv does work: and just for sanity, it does work on 17.03 and usually works otherwise O:) System: (NixOS: , Ubuntu/Fedora: , ...) Nix version: (run ) Nixpkgs version: (run ) Sandboxing enabled: (run )\ncc and since they were tracking down a similar problem in the Darwin clang stdenv recently.\nThis only happens on linux, it looks like the -isystem flags from gcc used to leak into the clang stdenv somehow.\nThis works, we could fix the clang stdenv to include those by default. Unless there's a reason to use the gcc headers, but using libcxx makes more sense I would think.\ncc\nIs this significantly different than using ? IMO it's an important use case, using clang w/gcc's libstdc++, it's the most \"compatible\" with Linux things and is the default way to use clang (for now, anyway).\nThis have something to do with ?\nFrom the name I wouldn't expect to still use the gcc libstc++, but the use case makes sense. The condition for  looks good, not sure why it's not working.\nnoticed that the logic would only apply if that variable was undefined, not empty. But now ensures that it is defined (and empty if not otherwise defined). The best solution would be to handle the default in so there's no empty fallback at any point.\nI'll do this for 17.09\nappears to be a victim of this issue as well\nseems to be defined before. I currently test the following patch: this is not a good fix though, but if this the root cause, I can come up with a better patch.\nThat would break or make it behave incorrectly.\ni know\nIs this supposed to work now? At least I am having the same problem on unstable(18.03); g++ works but clang fails and does not find the headers. I might have missed something though, rather new to nix.\nMaybe you need instead of ? Edit: nevermind, I can also reproduce your issue\nAfter some testing, I found that using without will cause the c++ header not found error. (I guess this is not a bug?) To successfully compile a simple file, I need\nSee\nIndeed it works in nix-shell, thanks.\nThere's probably a bigger issue here of people wanting our tools to work normally outside of Nix builds and shells, but that's a much bigger endeavor.\nYeah we'd need like a , which would easiest to accomplish if we did less dependency stuff in bash."}
{"_id":"q-en-nixpkgs-56fe156d4613c90c2263f76b39829a77e7a7d186ad3c5025fbd87dcff1e42339","text":"Package name: digikam Latest released version: 8.2.0 Current version on the unstable channel: 8.1.0 Current version on the stable/release channel: 8.1.0 Notify maintainers Changelog: Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-da4872123a651b1ab84aea54527e6c398a37d7d8d2ccec9110b07be6c574a6fc","text":"Seems like the wrapper in the installPhase is broken. When I build an individual core, like e.g. , the wrapper contains an additional argument that is always \"installPhase\". This is interpreted by retroarch as if you want to open a rom file called \"installPhase\", so you effectively are unable to supply a rom file via the arguments. Steps to reproduce the behavior: Results in retroarch trying to load a content file called \"installPhase\". Should run retroarch with core bsnes-mercury-accuracy loading the given rom file. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-9a69b90477f7e5095113d84588cbfa040caf96edd0108ed9368372d811c07c95","text":"nixpkgs input: cc\nsince this is broken to a degree that it puts a system in a state where it cannot be updated if any user is removed and no one has addressed this in the two months since i brought it up i think it might be a good idea to simply revert the PR - any thoughts on that?\nCan we fix this with a three line change or not? Can we check if the user exists in and then skip the disable? If the fix is not trivial, we should at least mark it properly.\nWouldn't a quick and acceptable fix be to not fail when disabling lingering fails because the user doesn't exist, and still clean up the dangling linger file? It does seem like a preferable fix to wholesale reverting this, and one that should hopefully be relatively minor.\ngreat so is someone able to take this on?\nThis is hopefully a quick fix, but I don't have the brainwidth right now to implement and properly test it, and likely won't until sometime next week at the earliest. I know there's been some work on a more comprehensive overhaul over in ; I don't know what that's waiting for to be mergeable, but unless it's ready to go now, my preference would be to simply roll back to stop causing problems for people, then roll forward with that change later.\n- thanks for the reply, i appreciate just hearing back from you on the issue - don't worry about scheduling, if you're busy you're busy :heart: if anyone has capacity to look at this, awesome, and if not i'll see what i can do... but i also would take me a few weeks to get around to this i appreciate that there is engagement in this thread, thanks everyone!\nI think I can fix this if I understand it correctly, which is that it is failing to disable linger for a user that no longer exists on the system.\nthat's correct i recently had to remove a user from a number of systems and if i recall correctly the systems in which the user had logged into, thereby activating lingering, would no longer activate until the lingering file under had been removed i would greatly appreciate if you fixed this thank you!\nI've opened to fix this.\nthank you so much everyone!"}
{"_id":"q-en-nixpkgs-0611c43b498a5a79a5455312cfef0eee8e64f9a7d7b212b800641a947b138356","text":"qt4 fails during the install phase ] The file in question is built but the install phase seems to be depending on the nix-store location before it is copied there. There also seem to be a few possibly erroneous file deletions by the build system: System: (MacOS 10.11) Nix version: () Nixpkgs version: () Sandboxing enabled: ()\nNot sure if it's something that we would want committed to the repo, but overriding the installPhase and manually copying bin, include, etc does result in working output.\nI missed the mention somehow, though I can't see how the issue is related to me. I can't say I understand the root of this. It tries to install/copy file(s) into a directory that wasn't created (yet). Perhaps some before installation would fix it? (I don't even have a Darwin machine.)\nNo worries; I edited you in way after creating this issue, since you were mentioned in the build definition (in case you had seen a similar issue on Linux). The installPhase seems to be deleting the framework (QtCore.framework/Contents/QtCore) before it gets copied, based on the fact that it is left on the system (nix-build ... -K) when the installPhase is shimmed. Now that I think about it I can probably add a postConfigure hook to remove the offending line.\n] ] Here I overrode the installPhase with It generates the library directories in $out/lib, but doesn't copy contents over as seen in the tree output.\nGives working output. I'm testing with a few qt-dependant packages and will submit a pr later today."}
{"_id":"q-en-nixpkgs-ee565f7709426f0fe917d312ba376636db30740f1bfc19528dc351e612344338","text":"With the standard library, cubical and categories, agdaWithPackages has a closure size of about 4.9 GiB (): Among the larger culprits are 600 MiB of libHSAgda-[...].a and GHC itself. Personally I'm using where I zero out the dependence on GHC and just fetch the Agda binary. I've been using this for years and never encountered any problems. Should we incorporate something like this in nixpkgs?\nI currently have a number of parallel installations of Agda on my machines, and I'd like to see a reduction of closure size. I think there are two separate space hogs: Haskell library. The mentions two helpers to reduce closure size: (removes references to libraries and docs) and (installs binaries in a separate output). I tried setting , but it breaks .<supSee below</supattribute passed to : I don't use Agda's GHC backend, so I provide a \"fake GHC\" to reduce closure size: Maybe could be changed so that disables compilation via GCH. E.g. we could pass and to the wrapped binary. I applied to the Haskell package . Running results in the following output: Result of run on x86_64-linux <details<summary2 packages marked as broken and skipped:</summary<ul<licedille</li<liemacsPackages.cedille</li</ul</details<details<summary1 package failed to build:</summary<ul<liemacsPackages.agda2-mode</li</ul</details<details<summaryBuild log of <codeemacsPackages.agda2-mode</code</summary</details"}
{"_id":"q-en-nixpkgs-ce0c8b03538b4c271ccb1da88af9b9407ab36015a587dea29313a39d34942235","text":"I'm messing around with the hostapd module a little & noticed that the value of in the config file was my SSID. However, online resources ( ) seem to suggest that this should take a bool (0/1) rather than a string. This is the link to the current configuration in the hostapd module: Hostapd doesn't seem to give errors regarding this, but I don't think it is correctly implemented right now. Could someone else take a quick look at this? I'm fairly inexperienced with hostapd and would love a second opinion / pair of eyes. If I'm correct, the only fix we'd need is changing to Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nYou are absolutely right, looking at the initial draft of the module it looks like this has been the victim of a refactor later :P Thanks for catching this, I've already opened a PR to address this."}
{"_id":"q-en-nixpkgs-42d35284993307bd797cc4a91a361cf589f0c15138793945bbc6e3fbfd5b8f06","text":"Package name: uvicorn Latest released version: 0.27.1 Current version on the unstable channel: 0.24.0.post1 Current version on the stable/release channel: 0.23.2 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThis should be closed, since the PR was merged."}
{"_id":"q-en-nixpkgs-bfdb9150b211705d4332b20aa249c98eeaa9d21497c2cb8490cdbc050e81be15","text":"Build of failed on x86_64-linux as of . This is currently breaking . maintainers cc: maintainers cc: Other cc: n/a nixpkgs at commit Contents of : This issue was automatically generated by . that created this issue. Internal tags: -\nI could not reproduce this, when I built the package it worked. Could it perhaps be some impurities in those tests? It seems it is some kind of port conflict (which the test seems to test). Seems those tests have been reported to be flakey upstream a few times:"}
{"_id":"q-en-nixpkgs-25d20ff64a2d9850805c2e422ef791ec1a1da56d49cba3b1b4ed0b92c33eda6a","text":"XPDF updated their version from 4.04 → 4.05 and seem to have broken the download URL to the older version. returns a 404, but works"}
{"_id":"q-en-nixpkgs-c47dc1cbcd3e0d0ad436f531478f1236d7dc49a8bdf8c0051d9ba11eab367b34","text":"A frontend in GTK 4 and Libadwaita for Apx. GUI for tool for (both are already packaged in nixpkgs) Metadata homepage URL: source URL: license: GPL-3.0 platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-eeb330d2df4b3889c916a04cb26d2f04903e74214edddbb09515e4da80c20be3","text":"Project description see Metadata homepage URL: source URL: license: unfree platforms: darwin, linux\nI was just checking myself and it's been there since November apparently\nAh, don't mind me, not checking the platforms..."}
{"_id":"q-en-nixpkgs-3d7ab574308d5772705ac70715f70fed55ac4e5421589c6cf8f4ab7db12861d1","text":"Package name: tmuxinator Latest released version: 3.1.2Current version on the unstable channel: 3.0.5 Current version on the stable/release channel: 3.0.5 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-3302d5373d414866c1b33ab0fca3d17fd001e203e02d8139951d6a9957f8a34c","text":"Package name: lightgbm Latest released version: 4.3.0 <!-- Search your package here: --- Current version on the unstable channel: 4.1.0 Current version on the stable/release channel: 4.1.0 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Version 4.1.0 gets broken by . Version 4.3.0 incorporates which builds tests with C++14. Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks. I'll have a look"}
{"_id":"q-en-nixpkgs-752088a0d7052a0d6b17668a797dc8557253494be4a1d120762630e64b143052","text":"I want to see systemd services in netdata monitoring. For that \"cgroup accounting\" is needed. The question is if enabling it on all systems would have any negative effect. In that case i would activate it only when netdata service is enabled. Arch Linux, Ubuntu and Fedora seem to have it enabled by default. debian not. shows no resources for services. System: NixOS 17.03.1700. (Gorilla) Nix version: nix-env (Nix) 1.11.13 Nixpkgs version: \"17.03.1700.\" Sandboxing enabled: build-use-sandbox = true\nI will implement it as option, that is enabled by default and for netdata service.\nI believe you can set this with: And maybe also: (ideas from )\nsure, that is how i do it right now. But i want to propose it to be enabled by default, so you have it when you need it for debugging. But also a single option to disable it when you have reasons to.\nIs this still open for the discussion about defaulting to or is this still open simply because it wasn't closed? Unless something else than the PR is required? (\nI created this issue to have this enabled for netdata. So an additional PR needs to be created to enable the new option when netdata is enabled. In addition it could be discussed to enable the option by default when a desktopManager is enabled (for desktop users). It should not be enabled by default on servers because it could have a little impact on performance.\nNowadays, this might already be enabled per default half-way: from my `systemd- (latest master right now) Are there any recent figures about performance implications when enabling the rest? If Fedora enables it, we could probably do so too.\nthis might already be enabled per default half-way thanks for noticing! can you create an issue or PR to change ? Are there any recent figures about performance implications when enabling the rest? i don't know any. i have no issues on my servers where i set it manually"}
{"_id":"q-en-nixpkgs-3fb0de96b929217bb9abe8c71d5d63a813ab7f0f567319edc9efd00a8ff91fbd","text":"Project description A fast, reliable and cross-platform command-line Minecraft launcher Metadata homepage URL: source URL: license: GPL-3.0 platforms: linux, windows (I can't test darwin) Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-73c2bebff6575167ca0f24747c9451bd3fb8a2d2dc3995593e95f8378fce18f2","text":"I noticed that in this line code actually checks true value of not that is declared at the top. Because of this, dlib doesn’t find the cuDNN libraries and compile without CUDA support. Steps to reproduce the behavior: with and without logs of both dlib compilation. One with will tell that it found cuda and cuDNN and dlib will use cuda, other will fail to find cuDNN and will tell that CUDA installation maybe incomplete and disables CUDA support. When the cudaSupport option defined in Dlib is enabled, dlib can be compiled with CUDA support without setting to true Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nAh good spot, I'll try and put in a PR to fix that tonight. Thanks\nAlso, I believe this issue can be related to this issue. For people affected by this they can enable = true; globally (which causes a bit of mass rebuilds on system). Also, I’m wondering why options defined in are not searchable in which could save me a bit of time.\nThank you too!\nLeaving this here for history:"}
{"_id":"q-en-nixpkgs-2d47c24a028fd083119d64ae7827e16ab4078bb257d03acea20e21ab21fb42fa","text":"Project description UI for building, configuring and running Monado, the open source OpenXR runtime. Metadata homepage URL: source URL: license: agpl3 platforms: unix, linux, darwin Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nTo my information, software without a stable release isnt accepted yet.\nNot even in unstable? Fair enough, I guess.\nunstable automatically becomes stable, so yes :) I think once a release is done (even 0.x), it can be packaged, but don't take my word as the ultimate truth, please :sweat_smile:\nThis appears to be incorrect. was merged targeting a specific commit (despite actually having stable releases).\nThis is indeed incorrect; there are even specific rules for the attribute of unreleased versions."}
{"_id":"q-en-nixpkgs-aaabb3def6eb30ebaf0248f9e03ab2f94e505cc9e36f63eec88653c1d0eacac3","text":"Steps to reproduce the behavior: (currently on staging-next) Upstream provides Python 3.12 wheels on PyPi. I need this for home-assistants Python 3.12 migration by end of March 2024 at the latest. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Issue exists on Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nWe can patch to use Also the linked issue at is fixed since 4 years.\nThanks for the sleuthing. Fix incoming.\nJust did the same in the background and opened"}
{"_id":"q-en-nixpkgs-18c8891dc808b69dc00bd7b1077d04c096210ccb8e6d1bdd546ac22c625a14ca","text":"Project description Scalene is a high-performance CPU, GPU and memory profiler for Python that does a number of things that other Python profilers do not and cannot do. It runs orders of magnitude faster than many other profilers while delivering far more detailed information. It is also incorporates AI-powered optimization suggestions. Metadata homepage URL: source URL: license: Apache 2 platforms: unix, linux, darwin, windows Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-4af1c3abfc6537cf9409ffef4d429d080e8d75f1e4ecbb55e03fae691f36bc07","text":"The kernel modules are missing which results in the lack of backlight control on devices such as the , GPD Win 1 and other Tables. Steps to reproduce the behavior: an affected device (GPD Pocket 1) i915 message: to control backlight, it doesn't work Backlight control works. Add any other context about the problem here. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nWas watching - can confirm pulling my kernel from unstable-nixos-small now includes pwm, pwm-crc, pwm-lpss etc in More importantly - now the backlight brightness adjustment finally works on my T100HA!\nGlad I wasn't the only one profiting from this change. ;)"}
{"_id":"q-en-nixpkgs-66d917816dac9890b4a684430a7912bade1a7a3d965275952dfb03f4755ec88f","text":"Steps to reproduce the behavior: spotify-tui-0.25.0 I see the package is basically abandoned upstream, should we just remove it? <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI think removing it is a good idea, give a shot :)"}
{"_id":"q-en-nixpkgs-5226c0c4290d4f6cfef1e41f18e8be301c7a6fda5c70a5b8b3d5499b6b014137","text":"Package name: gmic Latest released version: 3.3.5 Current version on the unstable channel: 3.3.1 Current version on the stable/release channel: 3.3.1 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-72ec1e00f94bd2c6cfc7272be1222af6888a8cc0c85efa5d1ad4efeb6a0a2d62","text":"Steps to reproduce the behavior: from Looks like a problem with building . May be related to . Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-42aecd945668e254756698364f80f0c030eae457b872693e00b2c458f04b9932","text":"Build of failed on x86_64-linux as of . This is currently breaking . python3Packages.torchvision maintainers cc: magma maintainers cc: Other cc: n/a nixpkgs at commit Contents of : This issue was automatically generated by . that created this issue. Internal tags: - -\nWill be closed by"}
{"_id":"q-en-nixpkgs-623b29bf12384dd5b7548100461293075468f3fa9fa6ed875fe7bfae5bcd62c2","text":"Package name: copyq Latest released version: Current version on the unstable channel: 7.1.0 Current version on the stable/release channel: 7.1.0 Notify maintainers"}
{"_id":"q-en-nixpkgs-001988e1d767dd9656ce063b6c896d7b3ff98641563f5cecac89937a271a98c6","text":"Steps to reproduce the behavior: . . Also failed in . <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI was experiencing the same issue regardless of the nixpkgs-branch I'm using for fprintd. I've asked in the matrix channel and came up with this overlay to skip the broken (timeout) test for the moment: ~(currently testing)~\nWorked for me after maybe because of some dirty state from previous attempts.\nAddressed the pam rule issue with\nHaving the same error building from this pull request.\nHow did you apply it? After applying it to my local-checkout of nixpkgs which I use to build my system from and updating flake inputs I was able to import (inputs-definition) as second package-set and set and was able to rebuild without applying above overlay. So this does not solve the issue of timing-out tests but it solves the issue that using the stable-package (which is the same version) wasn't helping either.\nWith merged, there is no longer a need for using an overlay to apply the workaroud, you may just set\nI'm also having the timeout issue, even with the workaround\nIf you're using the unstable channel, the still seems to be the best option until lands in unstable as otherwise it would require modifying the PAM module in addition to fprintd.\nUnfortunately, this still fails with timeout errors\nYou have to wait til it reaches your channel. See\nIf you don't want to wait and don't actively use a fingerprint sensor has been working for me without any noticeable drawbacks."}
{"_id":"q-en-nixpkgs-4ec2c873f7802f479d48c455fb755af2ba3a16859b2080bcd56a632f19d14a8a","text":"Anki fails to run with error . It was giving this error so I decided to update my system to latest nixpkgs, and the error remained. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nAre you mixing different versions of nixpkgs?\nI don't believe so. If you want to take a look at my configuration (it's a bit of a mess), it's . Oh, and I should have tested this before, but I just checked and works just fine, this is specific to .\nI tried installing and get this error too.\nI am updating my system to latest Nixpkgs. And going to check it once done. Edit: bty works just fine on my system, which is last update at 26 February.\nunable to replicate when building anki-bin on nixpkgs master ()\nOk I think I found the culprit. I ran under plasma6 and got the same incompatible QT library error as the OP. Then installed and booted into Pantheon instead where it runs fine. So seems to be plasma6 specific.\nah, makes sense. i'm running plasma5 still.\nNice catch. Then I would suppose it will be fixed once is released. As then ankitects/anki should be rolled out.\nif the bin variant is vendoring qt6, then we probably need to unset some envs otherwise things will often break.\nAnki itself does vendor qt6 version. But the package doesn't.\nForgot to mention that I'm running on X.org with i3\nI ran into it after updating my system to the latest nixpkgs, despite I don't run Plasma as well. You were correct. Following your comment, I managed to fix it by unset under arguments. Is there any further envs, that you recommend unsetting to prevent future breaks? Thanks!"}
{"_id":"q-en-nixpkgs-a1228ad6d768e9cd4886442828d7925ade8f75eaf1b530616cb6192875a3c84d","text":"The arrow keys, and also Home, End etc. don't work in konsole or yakuake. Pressing them does nothing - in particular it does not print an escape code, which makes me think that this isn't a problem. I am using these from XFCE, rather than KDE, but I would hope that that would be supported. Install konsole. Try to use the arrow keys. System: 18.03pre116572. (Impala) Nix version: 1.11.15 Nixpkgs version: 18.03pre116572. Sandboxing enabled: yes\nI think it was that mentioned this on the ~mailing list~ #nixos as well. Did you find a solution?\nI didn't, I switched to Plasma.\nPossibly relevant: running I get output for arrow keys, but nothing when running .\ni switched to xfce4-terminal but we need to fix this, konsole is rather important for us.\nI also switched to Plasma, where it works. One thing I think I discovered when trying to debug this is that it is not a issue, or anything relating to interpretation of escape sequences. Rather than getting uninterpreted escape sequences, it appears that konsole just doesn't get sent the sequences at all.\nxev started from konsole can still get cursor keys, konsole still can't receive them.\nThe navigation keys are not working because konsole can't find the data file for the . As a temporary workaround, I manually placed in my home directory under . For a real fix, konsole would need to look at the correct path in the nix store. Right now it relies on . Maybe this should include ?\nthis solution is awesome, can you also fix the problem in nixpkgs? afterwards contact me to get a reward!"}
{"_id":"q-en-nixpkgs-1dbdf57e1ac66cb6facafac77b9c8c6bfbbf72fb18d61753cdcd280ce908c4cf","text":"Steps to reproduce the behavior: Also fails on Hydra: I think the issue might be an old version of cython? Hydra lists python3.11-cython-0.29.36 which seems to be before Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nOverriding the package to use instead of seems to do the trick, but it breaks the pytestCheckPhase. I put this in my overlay:"}
{"_id":"q-en-nixpkgs-71d54f45ad1600261d1b8a6f95cd3a486cda57f9ced3ab78cecb9735e6ff3330","text":"Steps to reproduce the behavior: onthespot Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-a2b36a1ef9dd284ac874b215fcbe75a46de3922a3b877d359cb1fa3dd42c5c91","text":"Launching gparted on Plasma 6 wayland doesn't work Steps to reproduce the behavior: to (in my case I have it in ) GParted from Plasma 6 Wayland is promped for elevated privileges. Enter it happens, the window closes When launching from the terminal, this is printed after entering the password: GParted opens (through XWayland I guess) Other XWayland applications and Wayland applications are both working, through QT and GTK. The package has no maintainers :( last touched by so I guess I should notify Please run and paste the result.\nI found Arch BBS thread that might be related. In it, it says to put in your environment. Does that work?\nSame error in sway. Setting GDK_BACKEND=x11 does not help\nSame thing on COSMIC: system: host os: multi-user?: sandbox: version: nixpkgs:\nI've opened that should fix this, can someone test to see if the PR works for you?"}
{"_id":"q-en-nixpkgs-1defe19d55c7f5b901d97f7cd9e0592dbdaaa908936c439fa6964fbafc64686d","text":"On an apple silicon: $ nix-env -iA I thought was supposed to add support for macos. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result.\nWhy don't you ping\nI don't own a mac currently and I don't use the package, can't help you unfortunately."}
{"_id":"q-en-nixpkgs-64fda9bb684b02496eadfe5e5ef62e149d83ad838f626f7b484ad8b2bddd0b28","text":"Steps to reproduce the behavior: gnugrep-3.11-riscv64-linux stack-overflow-test failed <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-3395469ad48b09d2742480efb20a0d6cfba4541f2762ef727721646a7cadc5ef","text":"Steps to reproduce the behavior: on x8664-darwin Package is a dependency of . Builds fine on linux and aarch64-darwin, just fails on x86_64-darwin. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nif you submit a patch, I will review it, but I don’t own any Darwin hardware to test…\nPinging to see if anyone can help on this issue as it's outside my area of expertise. Otherwise will mark which depends on this as broken for darwin x86_64. Many thanks!\nCan test later, but it looks like there is a workaround for this exact error at\nThanks. I tried that flag earlier, which turns the error into a warning and allows to build (with warnings). But then when I try and build which depends on this I get linker errors,\nand were in 11.0. It should use to override the stdenv to use the 11.0 SDK with an 11.0 deployment target. Edit: It uses , so probably something like (untested):\nMany thanks, that worked! Much appreciated."}
{"_id":"q-en-nixpkgs-55efd96de5f9b1cd4795fbad2a9eae8b6e421e5e5ede9152d3dbb745637dbecc","text":"When trying to run , I am with the following Qt error: I'm fairly new to nix in general, and definitely not familiar enough with the Qt wrapping mechanism. I would expect the derivation to somehow set the \"Qt plugin path\" (?) properly so that is found upon startup of the application. I've found the section in the NixOS Qt documentation page which recommends using the in your derivation, but as far as I can tell, the derivation ? A dirty workaround I've found is to manually specify as follows, after which the application can be successfully launched: I've looked for a way to patch this properly into 's derivation, but I'm not comfortable enough with nix to figure out how to do this. I've found that other packages such as QGIS do : but it seemed to have no effect in , maybe because the derivation uses , and not ? Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nPlease test\nI merged the above PR. please give it a try once it hits the channel you follow:\nIt works like a charm, many thanks for your fix :heart:"}
{"_id":"q-en-nixpkgs-e1051ef78df56268c13e1c51ad1433661ec7e2a30eec69b9ebbf871a0b1bb665","text":"Package name: mesa Latest released version: <!-- Search your package here: --- Current version on the unstable channel: 24.0.3 Current version on the stable/release channel: 23.1.9 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-43062a6871d4493419803a94e60ef2174c65a1dbcd5024a75714082a44bed4b8","text":"Project description ClickUp is a project management solution, similar to Trello Metadata homepage URL: source URL: N/A license: unfree platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nSeems like it's needs to be archived with (currently it's down). Download URL is:\nI also found this PR, but it has been a year\nYeah I did give packaging it a shot but I could not find a reliable way to get it downloaded because there doesn't seem to be an exposed way to get a specific version?\nYeah, I also checked PR URL, and it doesn't seem to give reliable version package. Even AUR has some problems. That's why I suggested"}
{"_id":"q-en-nixpkgs-ad9bcddb5730e7b726ed9251f74700e326ce9ac985cfa14724bc0b1dfbfd2475","text":"Package name: sweethome3d Latest released version: 7.3 Current version on the unstable channel: 7.2 Current version on the stable/release channel: 7.2 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-311633c2522fe2194c160b8d9fbd5f0e1a8c72dfdf0d29b35b59202c61b21f42","text":"The last fzf update in nixos-unstable breaks certain bash-completions, for example git. This happens since Then try bash completion for git. My looks like this <details</detailsThe regular bash completions for git should work. As a fix I found this old issue in fzf To test this, I tried simply replacing with an edited version that puts the fzf eval after the bash completions code (and before the alias declarations). This fixes the issue. However, I am not sure how to solve this cleanly in NixOS. The bash completion code is inserted using but I don't know how to enforce ordering so that the fzf eval happens after the completions. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks for the report, I started a PR super-seeded by fixing this issue. As what I found this issue was already there previously but as most people didn't had the full integration enabled it didn't triggered the issue.\nIs there any progress on this? Solution PRs seem to have been closed?\nThe fix already got merged and will be soon on unstable - otherwise you might cherry-pick/overlay it.\nCool, so we can close this presumably?\nI looked at this in master and now we're basically back to how the fzf module was previously, right? So then it should still be an issue when enabling .\nAlright, so now I'm just gonna try my luck with a PR. It definitely works for me and makes a difference."}
{"_id":"q-en-nixpkgs-66c41d800a71d5c6eafe370b822b4419ce83cf9b624cd862575a5e8beccf0332","text":"Attempt to build duckdb on macOS. See eg n/a Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks and !"}
{"_id":"q-en-nixpkgs-332c14ddea0aa772bf263d5c32ffa64d76e711ecc27c93d6ef465e49564c09fa","text":"Project description It's a new version of Grafana Agent in Flow mode, but based on the OpenTelemetry Collector instead of the custom internals it was previously built on. Metadata homepage URL: source URL: license: Apache 2.0 platforms: linux and darwin Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-add3cf66f193e28a49cccb6c340d48a03e40c21a50e441841fa6855f74a798bc","text":"Project description A Mir-based Wayland compositor like , but tiling instead of floating. ! More Mir testers, more better. Requesting this from myself so I don't forget about this. Needs a very new Mir, newer than we currently provide (depends on ). Metadata homepage URL: source URL: license: gpl3+ platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-30755cc408e3c2ed45cb9b912f14860fb135d6b59b47f3d0137eec9a30d11091","text":"Recently, after updating my HA instance that is running on NixOS, I get when I start HA and I cannot use the SmartThings integration. Steps to reproduce the behavior: HA on NixOS SmartThings set up to whatever version caused it (I haven't been able to determine what version it was that broke) I expect the SmartThings integration to successfully run and connect to the devices it should and be usable, instead I get an entirely non-functional integration. I do not think this is an HA wide issue after doing some Google searching with the issue, I think it is NixOS related, but someone can correct me if I'm wrong. Can't tell exactly who to ping here, but I think at least Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nPlease post a complete traceback if available, so we can determine what exactly requires .\nThank you!"}
{"_id":"q-en-nixpkgs-50392005866fd27b4218a98d24b4bdfa16be02907c6e792d026779dc41c9715d","text":"Steps to reproduce the behavior: ollama service the model dir Ollama should work, files and directories have rw permissions for anyone and the file system is not read-only. It used to work before an update. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nCould you try manually renaming the file to see if you get the same error? If you still get the error, the problem is probably with your system config. If you don't get the error, the problem could be with the ollama binary, or with the service config. Maybe try it as root as well to see if there's a difference. By the way, the reason it's trying to rename the files is due to a change in .\nRenaming files solves the problem of starting ollama, but does not solve the overall problem. I can't install a new model, for example.\nI found what the problem is. The ollama service has set, which creates a new user for the service and enables some sandboxing that includes making most of the filesystem read-only. I'm not sure at the moment the best way to fix this without compromising on security, so I'll do some more research on how systemd works, and any relevant systemd options. At worst, I should be able to create a nixos option that disables this sandboxing entirely. If you have any suggestions, I'd be happy to hear them.\nIn the meantime, a temporary workaround that I think should work is this nixos config:"}
{"_id":"q-en-nixpkgs-05b572cf5ce44e7102fb8187bc28804fb00553e7f627624d76a7ee2b460f1f63","text":"Steps to reproduce the behavior: kvmfr <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-1a1945fb69a170ea3bd2dcce04364e7c6f750d12394460803f25a5df2c5f7b10","text":"Steps to reproduce the behavior: x3270 on darwin fails due to missing macOs dependencies (libiconv/Security) The are a few things missing that are need to get this to work on macOS use & link iconv use Security package for headers disable x3270 build (the other unix builds are fine) Please run and paste the result. I have a fixed package for this that I can open a PR for"}
{"_id":"q-en-nixpkgs-53b37fda00f2c3b6f7ddc7c67a66eaa40def9f7650a5a9245e28cdb4d60ec6e3","text":"Package name: i3bar-river Latest released version: 0.1.7 <!-- Search your package here: --- Current version on the unstable channel: 0.1.6 Current version on the stable/release channel: 0.1.5 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI looked into this a couple days ago. The problem is that 0.1.7 makes use of a features introduced in Rust 1.77. NixOS isn't there yet, so I don't think this package can be updated just yet."}
{"_id":"q-en-nixpkgs-f0c83ff79c29cf669518dd61608a4dfeb91e0b0e9561bd3dd79a045f93cc7e76","text":"Project description ab-av1 is a VMAF-based reencoding frontend for ffmpeg, primarily for av1 but works well with x265 and other encoders. It is a rust project and requires ffmpeg newer than git-2022-02-24 with libsvtav1, libvmaf, libopus enabled. Metadata homepage URL: source URL: license: MIT platforms: unix, linux, darwin, windows Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-1f2302bdab8f0a6262771e54529072d67d6889fd2e0aa66f6dc2d94794911710","text":"A clear and concise description of what the bug is. The integration seem to require the missing python module and currently fails with this error message: is provided by python . Adding to in the . Maybe this missing dependency could be resolved in pyowm as well. I would appreciate some input from somebody with more experience. Steps to reproduce the behavior: home-assistant with openweathermap component No Error message and working openweathermap component <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nPR\nPlease post the full stacktrace.\nThere also was an upstream issue that may be related:\nThe fix is to add to pyowm's dependencies.\nok. Thanks for your help. I will create another PR later"}
{"_id":"q-en-nixpkgs-b5877da8c04649430d63261a213fbd04662ec8cd97a783daceb434a7a71faec7","text":"OAuth2 Proxy's Nginx plugin used to forward and headers to downstream programs. This stopped working after I updated on unstable channel, which is potentially related to or . Steps to reproduce the behavior: up OAuth2 Proxy + Nginx up of header name . does not recognize the session More simply: up OAuth2 Proxy + Nginx or something similar the headers — and used to be present, but not anymore and are passed to downstream servers under Nginx. I have I went on upgrade of nixpkgs unstable from sometime late January to a few hours ago, which is what led me to the two PRs mentioned above. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --cc Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nare you able to play around a bit and confirm if the changes mentioned introduced this regression?\nyes. I can't really pin-point to what exactly since there were a lot of changes. Instead, I just have a local fork from (prior to the changes mentioned in the PRs above).\nI have a pretty good idea what it could be.\nCan you give a try? I didn't test it yet, so it could completely break everything and be completely insecure, just as warning.\nfantastic work please let us know how it goes ‍\nSorry for the late reply, I haven't been able to have time to test it out. Now that I do, it seems like has been converted to . This won't be an easy find-and-replace for me to try in my setup, so I will have to find some time to do proper cleanup."}
{"_id":"q-en-nixpkgs-6ebf0aedc167d6dde9413e6ecfda6d0b5f4bccd1663ab9db083ba9d83b7e430e","text":"Package name: snapcast Latest released version: 0.28.0 <!-- Search your package here: --- Current version on the unstable channel: 0.27.0 Current version on the stable/release channel: 0.27.0 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-90c64f30339f2405e4ae636b08838dda82f84da84fbe6dc18cc01a076da969e0","text":"Package name: gummy Latest released version: 0.6.0 <!-- Search your package here: --- Current version on the unstable channel: 0.3 Current version on the stable/release channel: 0.3 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- There is an open pull request () by from October 2023, but it has not been merged and later versions of the program have been released since then. I think it is also worth noting that the source code repository of the is no longer available on GitHub. as the repository has moved to Codeberg (see ). Notify maintainers (last updated the package) <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI updated my linked PR to the latest version.\nThank you for updating it so quickly! Is there a time frame in which we can expect the pull request to be merged by someone?"}
{"_id":"q-en-nixpkgs-bb0269c1bd720eecac63c4261fc8a540cd26a3428b10f10b66bcf9e8dc917fe9","text":"Package name: clapper Latest released version: 0.6.0 Current version on the unstable channel: 0.5.2 Current version on the stable/release channel: 0.5.2 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-1a297d491ab320066101d0172105ecb1326a4acedf7fe4624969084bf1ec8cc9","text":"example: This one is low on the massive bucket list, but extending the number of retries and the timeout between each retry will reduce friction.\nThis might be caused by , in which case increasing the timeout wouldn't help much. We can try though, PR appreciated!\nplease view"}
{"_id":"q-en-nixpkgs-44bb68bc7763032f32d168b31b4e2dffa2c1dcd331cd78af00bebcd9f6a04007","text":"When installing usbimager the path in the desktop file is /usr/bin/usbimager, so when you try to open the application it doesn't open, to fix it you have to copy the desktop file from /run/current-system/sw/share/applications/usbimager.desktop to and edit the desktop so it reads Exec=usbimager Install usbimager, try to open the application from your menu, at least this is the behaviour on xfce, I've not tested it on other desktop environments system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nDoes it work in a nix-shell? nix-shell -p usbimager usbimager Worked perfectly for me on hyprland.\nIt works in a shell, it also works when installed system wide when you launch it from a terminal, its when you try and launch it from a menu like xfce's whiskermenu it doesn't launch, because the desktop file has Exec=/usr/bin/usbimager"}
{"_id":"q-en-nixpkgs-da39cd7469b63284ba9b4c8d3a3de0871d989693ef7d19184c02a9b9ac12c1c4","text":"Koodo reader is an electron application, and although the build process to add the proper ozone flags when the app is launched in a Wayland environment, it seems for some reason, it's not working as expected: But launching the app with the required flags, it works: The process arguments seem weird: Notice that the the by the package are truncated after the first space . I tried splitting up the flag into two sentences to avoid spaces: But now it looks l like the process does not pick up the environment variables: The variables exist, I define them as . Changing the above override to explicitly set the ozone flags (without the shell expansion conditional) works. Other packages like do the same shell expansion, and they also work in my environment without overriding anything. Steps to reproduce the behavior: in a wayland environment (I'm using hyprland) if the application is using xwayland by running shows the application again but with the ozone flags the app does not use xwayland: Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI'm not a wayland user, so I usually just copy this line from other electron apps' definitions and hope it works. I think the issue here is that when is present will use the binary implementation, which doesn't expand shell variables, thus not making this work. It looks like I forgot about this behaviour when packaging this."}
{"_id":"q-en-nixpkgs-92efc236736aa20c09b22df2313abb633f1f0f8176e1b8a09fdcddcb6a41d0e0","text":"Package name: ms-vscode.cpptools Latest released version: 1.20.2 <!-- Search your package here: --- Current version on the unstable channel: 1.17.3 Current version on the stable/release channel: 1.17.3 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-e4817bb1803ca4e20ca8ce02750312474b7ba2521bcf01dcb2d4c452bafe6270","text":"Steps to reproduce the behavior: python311Packages.dulwich Package recently updated and maintainer used the wrong source tarball path Used but should be Please run and paste the result. system: host os: multi-user?: sandbox: version: nixpkgs: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-b624e2e8c4cee3c185a7b6eafba4a27ad8422bef0d37215a6d958e57376a11e2","text":"No crash n/a n/a Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nSame problem, same message, with the following system. WM: sway\nAttempted hacky workaround from This changes error to Progress?"}
{"_id":"q-en-nixpkgs-a0d77f17df829f9cd4ee8b6c20e8434aafec46bfc0d019efbf818fd0b8556a3e","text":"A clear and concise description of what the bug is. In , XDGSESSIONTYPE is now set to wayland by default for Hyprland users. However, SDDM will crash when this env var is set. Steps to reproduce the behavior: SDDM with Set under . SDDM greet will never show, and the screen will be black. These two lines can be seen in journalctl, and they are not present for boots from older generations. SDDM does not crash. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nDid you enable wayland support for SDDM? ()\nThanks, enabling that option fixed the issue! Would be nice if there is an config assertion for this.\nHaven't checked the implications but maybe sddm should just enable by default?\nenabling wayland for SDDM, makes it impossible for me to start X11 only sessions (eg Pantheon) Is this expected, or is there a way to use SDDM with wayland and X11 sessions?\nSame issue here, workaround mentioned in worked for me. I found the following stacktrace in the logs, not sure if/ how this might be helpful.\nThe proper workaround for now would be to set . Those are envvars that should be set only when the hyprland session starts, not categorically set.\nThanks, this worked for me. I can now start X11 sessions again :)\nSame problem here using Hyprland with Budgie as a second desktop. Thanks for solution :+1: works!"}
{"_id":"q-en-nixpkgs-2ad151effaeea91f6f10ce3eeb9184b52fd4392437b6c9805fe88b556d231412","text":"Package name: LLVM Latest released version: <!-- Search your package here: --- Current version on the unstable channel: 18.1.4 Current version on the stable/release channel: 18.1.4 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nWant to submit a PR? The last upgrade was very straightforward:\nI'll give it a shot"}
{"_id":"q-en-nixpkgs-95da61596604ab7e0ca69a5883f52f333767cffa6bf7f6d032a0fb34c4d67d1d","text":"For Intel dGPUs, requires in order to access GPU utilization. Steps to reproduce the behavior: an Intel dGPU, such as the Arc A380 and the following relevant logs: This error should not happen, and should be able to read GPU utilization. ! As you can see, GPU utilization information is not being displayed. N.A. CC Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nHi It looks like there's an upstream issue tracking this: I'll look into this more to see how we can get this working, but I don't think this is something we can solve directly in the mangohud package. We'll need to use a separate NixOS module for since it'll require a security wrapper to enable . After creating a intel-gpu-tools module that sets the capabilities, I'm pretty sure we'll be able to get it working through something like this: but this would need to be used in addition to installing mangohud. Installing mangohud by itself wouldn't be enough.\nWas able to get this working : Before | After ----------|--------- ! | ! Now just need to add this to a new module."}
{"_id":"q-en-nixpkgs-6ab002679b6758892955f2713c9d3e92deafb260b6567e9c3d2d54dbce883ecc","text":"Steps to reproduce the behavior: blackbox-terminal blackbox uses vte with , this change probably broke it Please run and paste the result. but that's irrelevant as i'm trying to build the latest master. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-f7cf1943bef58db9d48ef8cc828e614b8a5b0028dfc9dc12227c0fc6b8a2503c","text":"Project Description An actively maintained picom fork that adds extended animation support. Metadata homepage URL: source URL: license: mit, mpl-2.0 platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nOne minute, I need to push the right branch\nhas been done, please react onto the pr so that it is marked as being more important. Have a very nice day\nThanks a lot! Added the reaction to the PR too.\nIs this working? I have changed my picom configuration to use this package but none of the new options work.\nIt should be working, at least it does in my configuration, which includes the package before I ported it to nixpkgs. If you could state what problem you are having, that would be nice. Unfortunately I won't have much of a way to help you because the IC of my Linux laptop is fried, meaning that currently I have no way of being on Linux right now to test anything."}
{"_id":"q-en-nixpkgs-4639429f1ec11d83cd05f454edab1c0096d4ee81604e41452b5a7c8503f1ba3a","text":"Project description WebUI service for Large Language Models Video demonstration by : .. basically self-hostable ChatGPT-like service Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin Additional info Follow-up on for open-webui nixos module. There seems to be an ongoing work on the subject by and on Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nRelevant Merge Request:\nJust for reference / workaround here's another nix package:\nThanks for that reference. That pdm tool workaround looks a lot more maintainable for the Python backend. Maybe i'll have time to try another PR in the next weeks. In the meantime and as a much simpler UI i packaged as part of PR It would be great if someone could review it.\nI would like to mention that in the latest release switched to , I am not sure of any straightforward way to do this while packaging it for .\nI just tested your flake, it's impressive, especially the nice UI at the end showing all the processes ! Regarding open-webui, it would be so nice to have that in !!!\nThank you, planning to have more UIs and devShell integration (for CLI tools that interact with ollama). I would love that too! Although there are a few challenges (like the one I pointed above regarding ), I am willing to contribute here. Also, maintainers are quick to respond, in case we need any upstream contributions to make it happen."}
{"_id":"q-en-nixpkgs-6e335f270883fdf5669c0fe56db365902db211db7c4f73ee71a5540c80ec2e4c","text":"If you have: During evaluation I get this error , pointing to the line Steps to reproduce the behavior: As described above I think this is because here The pairs are made using the , while here: It is trying to use the vhostName. So I think it should be Also, maybe this could be a problem as well, however I am not sure: I would submit a MR, but embarrassingly, I couldn't figure out how to test a patched nixpkgs lib in a flake without having my whole system rebuilt. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nBroken by . Fixed by ."}
{"_id":"q-en-nixpkgs-a2635dc1301ba3345b03f75859e2607ac49588e3cb826925a8ed122aa9069191","text":"The patched version of Ardour in nixpkgs is not able to open VST2 audio plugins also installed from nixpkgs. All VST2 plugins are detected, but say they can’t be opened by Ardour. For example, having installed the ardour and lsp-plugins nixpkgs packages, the Ardour plugin manager shows the status “Error”: ! The error given for one of the VST2 plugins is: Steps to reproduce the behavior: Ardour and some VST2 plugins from nixpkgs (e.g. \"ardour\" and \"lsp-plugins\" packages) Ardour and scan for plugins The plugins are detected and can be used. I'm running NixOS 23.11 and installing these things declaratively with home-manager. i.e.: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI suspect is not actually a plugin, but a shared library used by the plugins in the same directory like which loads fine for me: Are you sure the actual plugins are affected? !\nI tested with and and I was able to see both of those, but I'm able to reproduce other vst2 plugins (like LSP) not showing. Could it be that they're not in ardour's vst2 scan path? !\nnixpkgs patches Ardour to find other plugin types, but I think i've to the plugin search path manually.\nHi guys, thanks for taking the time to look into this issue. You're right about the shared library not being a plugin - this was a red herring that was stopping me from understanding the problem. I just to my plugin search path and everything I was missing has now loaded in successfully. Thank you! To prevent other users from experiencing this problem (and possibly me if I change systems and forget this fix), can this path be to the list that the nixpkgs patch adds to ardour?"}
{"_id":"q-en-nixpkgs-e491ec71b16fcee0e1672cdd119b8e29ff32cac5b3d1c852c332c8d34b4f317d","text":"looking over broken python packages for ZHF and pixel-font-builder is missing a python module and needs a dependency relaxed. Steps to reproduce the behavior: *pixel-font-builder\" pcffont not part of nixpkgs would add it but no longer a maintainer <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-834da8b52f7c1b246f0aefa51f77d49c81f4d3ad6fa54ee1f17345d5e62a260b","text":"Project description Utility to activate iPhones Metadata homepage URL: source URL: license: lgpl-2.1, GPL-3.0 platforms: unix, linux, darwin, ... Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-102f8a6ecbc4b4a5f80d941d9b583d210165b3d8f50f9d6e9098a934851fe89e","text":"Open a C++ project in vscode with the extension installed and see the following in the Extension Host output window: Steps to reproduce the behavior: a C++ project in vscode with the extension installed how autocompletion etc. features do not activate. to the output window and select the Extension Host and you will see the following: The extension should work ! Add any other context about the problem here. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI'm wondering if that should be 0555 instead of 0444.\nAdding to got me around this particular issue but now I'm running into which should be resolved by moving to 1.20.5\nMoving to 1.20.5 doesn't fix the startup issue\nI'm replying to your comment here I can reproduce this with the following: Attached is the flake I end up with. My system packages are the following:\nAfter changing the permission of this is what I get: If instead I run through manually I make some progress (it's looking for in my home directory) and it exits.\nThis is most likely"}
{"_id":"q-en-nixpkgs-6c76985f9ba4ea0c629c29754607071b92b9f8a22412e0f3c34b389109269ab7","text":"The following example in the python docs does not work since on the second line is supposed to be the interpreter and is for the packages: The fix is simply: System: 18.03pre118631. Nix version: nix-env (Nix) 1.11.15 Nixpkgs version: 18.03pre118631. Sandboxing enabled: false"}
{"_id":"q-en-nixpkgs-f7744703e0b8595969d525ccc9fd3458d5a89071c19968267d0b4890cbeceb07","text":"When trying to launch whipper it instead fails with the following trace: Steps to reproduce the behavior: or any equivalent way of running the whipper program Clearly, I expect the program to execute without an error. I bisected the error and the bug was introduced in which bumped . <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nNote: Adding to double-wraps all executables, but allows to run successfully.\nAlso no double-wrapping."}
{"_id":"q-en-nixpkgs-cf1590e34d4a154423d7053aef365ad81f5b659850a5cc3085be10ed8e38defa","text":"a TUI-like CLI tool to manage the features of a Rust project Metadata homepage URL: source URL: license: mit platforms: everything that Rust supports Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-5d8ae951a2437dadc104ffbe6eec896c520d7616ef0c3a03564a491c1b9f14ab","text":"When configuring notifypush for nextcloud the service almost always fails the first time after rebooting the system and requires to be restarted in order to be started successfully. It seems the problem is that notifypush has not finished starting before nextcloud-notifypush-post-start is executed, resulting in a missing or not connected . Steps to reproduce the behavior: notifypush the system (un)lucky and watch nextcloud-notifypush.service` fail service (done automatically) The service starts successfully the first time. Relevant log lines: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI have noticed a similar behavior when the module it was introduced. But i was so far happy with the restart situation. I have played a bit with the situation. I am currently applying the following additional configs to my nextcloud. I think the problem is, that notify_push creates the used socket at start, and we are to early executing the ExecStartPost command to configure. When we let systemd take care for the socket, i could mitigate that and after that i had problems that my acme certificate reload was running in the mean time. After adding nginx-config-reload it was able to run on first try after reboot. You could check if that is also helping on your end and if yes we could check what make sense to use in the module.\nI am not sure if this is a good idea since existing Unix sockets are on server startup (and shutdown), possibly interfering with the socket management of SystemD. I don't know why the connection succeeds after restart or when the socket provided by SystemD is replaced by another socket by notifypush, but I think a more robust way to implement this is to set which would delay the execution of until the server signals that it is ready. However something like this would require support by upstream, I could work on a pull request. Edit: After doing some testing it seems that for notifying SystemD to work notify_push has to be the main process of the service, which can be arranged easily by using in the start script.\nDoesn't have the program call sdnotify for that to work? I did not find something like that on the code base. here the notify section claims something like that. Tho I fully agree that the socket is likely not a good idea when notifypush nukes the socket file. Edit: NVM you mentioned a upstream pr for such a function , which I seem to have over read when reading my mail notification."}
{"_id":"q-en-nixpkgs-0f5c4a55c6056a0ed42f7dc485e439229007f56a41f4d96c0eb0ad294dc12a32","text":"Hey all, not sure which template is most appropriate for this, apologies if the format is incorrect. It looks like this package no longer exists upstream--the repository owner seems to have deleted or renamed their account. I'm not aware of any maintained forks of this project, so I think this package is completely dead, and should be removed from . Note: while searching for forks, I noted that 1Password appears to . Without access to the original package, I can't say for sure they are equivalent, but may be worth investigating for any affected users. Package name: Latest released version: GONE <!-- Search your package here: --- Current version on the unstable channel: Current version on the stable/release channel: <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-72a5c03d70445d14eff66b5393993b0269e2fbd06c9ecf44277f5a4dcae32777","text":"A clear and concise description of what the bug is. Steps to reproduce the behavior: Mac Silicon, Utilizing with nix-darwin, i've the folllowing : always get qemu-system-aarch64 as qemu binary and then an aarch64 linux builder uname -a should give x86-64_linux Qemu binary program in () should not be forced to hostPlatform system for this linux builder case If applicable, add screenshots to help explain your problem. Add any other context about the problem here. Did not find any issues related to this qemu program choice.\nI think the qemu binary is set correctly in the builders, in this file: I was also having this problem, and found this issue while looking into it. But I realized the result is being caused by these lines in nix-darwin: () That looks fine, but the builder package we're trying to use is also created using override! So the in nix-darwin is overriding the in nixpkgs, making the resulting package have the default hostPlatform. You can work around this by setting that value in your nix darwin config, like this: I've opened a pull request to make this work the way we expected it to: Thank you for this issue, it confirmed that I wasn't the only person having this problem, and led me to find the cause.\nThanks a lot !"}
{"_id":"q-en-nixpkgs-9775df2e04e73a1df84654c654e105740ec019a1c87ca4b92477d93aef1d40d5","text":"Package name: Latest released version: <!-- Search your package here: --- Current version on the unstable channel: Current version on the stable/release channel: <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers /cc <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-7e7e099bf484636031e43e02cb4fb2682016a3d00e9437d00721ce63d64f7591","text":"Steps to reproduce the behavior: to the NixOS configuration. nixos-rebuild to build NixOS. According to the Nixpkgs search, there should be the packages python312Packages.journalwatch and python311Packages.journalwatch in the Nixpkgs unstable branch. I just checked the master branch and I haven't been able to find those. I could only find pkgs.journalwatch. That's fine, but then it means the journalwatch module needs to be updated. Perhaps this is related to <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --I won't add the Metadata since I'm in the process of building NixOS, so the result would be incorrect anyway. I'm on Nixpkgs commit Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nBreakage caused by . Please have a look at and let me know if that works.\nRan into this as well, the PR looks good and seems to fix the issue. Backport should land soon. Thank you all!"}
{"_id":"q-en-nixpkgs-777348a90017a07a7256e52c4e04bea8f69f252116fc335d27cb71f122cb95da","text":"Mainline / hardened / testing / ... kernels can be painlessly modified via : Sadly, this does not work with a bunch of other kernel variants: All kernels should honor I am not entirely sure how dynamic binding is implemented in kernel stuff, however slightly changing xanmod parts this way seems to fix : zen and lqx look similar to xanmod, no idea about libre Pinging people who seem to update affected packages often Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nsame underlying issue as it seems\nCan confirm that this patch fix the issue for zen and lqx kernels:\nhave you submitted these changes as a PR that I can review?\nsorry for the late response. I created a new PR with this changes.\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-ab0951e8cbc51f3685e46b30824e4652a7d73fa04cf39c3a24e695c8ef7281fe","text":"Unable to use Podman's quadlets feature for a user, since the service unit files are not generated by systemd. Steps to reproduce the behavior: a quadlet file in (say, ) . Systemd says that no unit files were found. Caddy service should have been generated and its status displayed. This issue is caused by not symlinking the provided by the package to . Does not occur at the system level, since is already symlinked to Fix: Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nConfirmed the quadlet systemd user generator is not found in /etc/systemd. I'm gonna take this one and try to come up with a fix in NixOS module."}
{"_id":"q-en-nixpkgs-bc01c0bc09a7578f153d7e6c949c7a5b2dd47ee6392d3c909f8924fcf4de07e9","text":"luarocks exists that other luarocks packages can depend on. When trying to build a rockspec that depends on luarocks with , it fails. The entry in the generated luarocks config contains an entry for luarocks, but luarocks doesn't recognise it and tries to download the package instead. I have created a repository to reproduce it with: Run A workaround is to use to generate a expression for luarocks (), and to add that package to the 's instead of the package from nixpkgs. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nthanks for creating the test repo, it's helpful. I've looked at the generated LUAROCKSCONFIG and all looks good and I think luarocks just doesn't recognize the installed luarocks because of My first thought would have been to rename the current luarocks into a luarocksbootstrap and add a generated derivation, for which we could run the tests without the cyclic dependency luarocks -busted -luarocks. While this might be the best outcome longterm, I am looking into the quickfix which would be to generate the manifest for the current luarocks"}
{"_id":"q-en-nixpkgs-6875edaf0b1db1e4f81cfb41bed0a7740edb4d2a8e902d0c9f2b60fa16753434","text":"Steps to reproduce the behavior: archivy Add any other context about the problem here. nixos 24.05, nixos-unstable at Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-d2d4031eb925a53f17db1909db2eda16ffd5714e4031dc50f69b6e25156018d0","text":"Project description Instrument your LLM app with decorators or low-level SDK and get detailed tracing/observability. Works with any LLM or framework <!-- Describe the project a little: --Metadata homepage URL: source URL: license: mit platforms: linux, darwin Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-6c76c0a0746891f0d47019f9714bfdfbbb6acc233b88bc4280ae4cb113eecb62","text":"Steps to reproduce the behavior: radicale at quick glance I think they rereleased tarballs with more changes, files inside are from today Please run and paste the result. system: host os: multi-user?: sandbox: version: nixpkgs: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nYeah, looks like upstream changed the release afterwards. I didn't even know you can do that on GitHub.\nYou can just delete the tag, create a new one with the same name and do a force push. Some projects do this, some don't.\nI suggest any that do be removed from Nixpkgs. It's unacceptable.\nCan someone revert for now?"}
{"_id":"q-en-nixpkgs-2b50520d2d6247c9842541a0d661a3616120ff4ccacd5d8d95e2ab64bb130d14","text":"Steps to reproduce the behavior: The src is: We tried to use from nixpkgs in , but that made nix-build . -A fail. I guess that happened due to one of the commits between and (which is the latest commit on ) <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-f88cc6f2bc97b2b6aae8bb0a1a6ec3e914fd384609401ec2618c135ffdfd4968","text":"Project description <!-- Describe the project a little: -- alternative for GNOME (I guess) Metadata homepage URL: source URL: license: gpl3Only platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-661b37b5dcd036231410db9478dd7470d4ed3ece1ecdcc0ecb89d9a410e8d051","text":"Sometime between 23.11 and now, the binary broke due to a linking issue. Steps to reproduce the behavior:\nfixes it of course, but I don't see an obvious less brutal approach.\nadding to the works for me.\nGood point, somehow thought this was already the case. PR up."}
{"_id":"q-en-nixpkgs-11e29c55b06fba99994639e3538d4ae156562b0af66fc24cff555691261c33c9","text":"(///) packaged in does not detect Muse Sounds Manager. The Muse Sounds profile is greyed out and cannot be used; leaving only MuseScore Basic available to use. Steps to reproduce the behavior: from (at the time of updating, version ) the unofficially packaged ) flake. MuseScore either from the desktop entry or via in a terminal, visit in the top bar, to check for available playback sounds to use. The Muse Sounds profile should be available to use and not greyed out. ! I have acknowledged the fact that although the mentioned Muse Sounds Manager package is unofficially packaged, the package works with the AppImage Version of MuseScore (packaged officially ), so I don't think this is the problem with the Muse Sounds Manager flake. Below is the screenshot of what the Playback Setup window should look like, taken from the AppImage version while opening the same score as the previous screenshot. ! I have tried reinstalling Muse Sounds Manager itself and reinstalling each Muse Sounds, but the issue persists. My entire system configuration files are . <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nSorry, miss-clicked submitting the issue without adding any information. It'll be updated soon and reopened later.\nHey I'm sorry but I can't reproduce your issue. I Ran and downloaded an example font, and with both 4.3.0 and 4.3.1 I get: ! Before I downloaded the sound fonts I also had that button grayed out. Maybe you simply need to check again?\nThank you for responding. The issue persists after I've tried and and everything else (reinstalling MuseScore, Muse Sounds Manager, reinstalling all combinations of Muse Sounds) again to confirm that this is not an isolated incident. Additionally, I've compared the logs of both the AppImage version and the version. Most of both outputs are identical (apart from some timestamp-dependent data), but these lines only show up on the version: If necessary, please let me know how I can provide any more information.\nNice! Perhaps it is related to specific sound files you downloaded with that I didn't? Perhaps can you share all those? I have only this one named .\nI've tried and downloaded all Muse Sounds from the beginning again for another time just for good measure. Here's an exact order of what I've tried (only one at a time, uninstalling all other Muse Sounds) and the issue still persists: Keys Brass Guitars Vol. 1 Choir Harp Percussion Strings Woodwinds So, all of the Muse Sounds available. The terminal logs print out the same error message on the version. I've also tried each of these with the AppImage version and I can confirm all of these are recognised with no issues. Is this in some ways related to how ? (This is an uneducated guess from me noticing logs mentioning the dynamic library )\nHey and sorry it took me a while to get back to this. It'd be nice to get your approval upon the 4.3.2 version bump (haven't even built it yet): So I tried to download all of the sounds mentioned above, and still I can see the muse sounds profile... I don't know what else I can suggest to debug this :/ Wait where do you see these lines? I don't see them in either muse-sounds-manager's logs nor in Musescore's ... Log of muse-sounds-manager-nix : And log of Musescore:\nIt's all good :) I'm just happy someone is even looking into this, to begin with. As to the 4.3.2 version bump, what do you need my approval of specifically? I'm sorry, I'm still very unfamiliar with the Nix ecosystem as a whole. Although, I have built the 4.3.2 on my machine and tried , everything seems to work (I've tried playing some tracks, editing them, etc), including the Muse Sounds. ! Additionally, I've done some more debugging: It seems like: running in the terminal (tested with + ) doing and then in the terminal (tested with & ) launching from the Desktop Entry all had the issue of Muse Sounds not being available to use. However, doing: (in ) / (in ) , into the output directory, then do all have Muse Sounds available to use. So this seems to be more of an issue with my shell environment somehow, and I honestly have no clue why this is the case, or where to start looking. That's a bummer. I mean, I guess I could always just use the AppImage version; It's not a big deal. Thank you for looking into it at least. Honestly, it could just be something arbitrary only in my machine in particular. Sorry, I should've clarified: The logs I sent earlier are from MuseScore (I just ran in the terminal). I didn't share Muse Sounds Manager's logs since nothing seems out of place there. Here's the full logs of me only running in the terminal and then closing the MuseScore window shortly afterwards: As you can see, line 9~12 showed the error I mentioned earlier. Again, thank you for taking your time looking into this.\nIndeed this line seems suspicious.. And when you run version 4.3.2 you don't see it? I didn't find that line in my logs as well...\nYes, all other mentioned methods with Muse Sounds available don't have those errors in the logs.\nIt's weird because returns nothing... There is: Which may be related.\nEdit: I was mistaken, setting the still doesn't detect Muse Sounds.\nYes it surprised you wrote initially it worked for you. We are on NixOS after all :). I'm not sure if you are subscribed to that issue or not, but I asked them about the source of .\nBTW today I experienced crashing for me, and I checked the backtrace () and this is what I found: Interestingly: So indeed this issue is related to external shared objects that are not compatible with NixOS. I'm still a bit confused as to why in your case it doesn't necessarily crash, and instead it silently fails and you don't see the muse sounds... It'd be interesting to try to that file before you run ... In anycase, this should also be discussed with .\nHello, I've done some additional testing. It's not much, but I still want to tell you what I've found. It turns out these only work with that one MuseScore instance. From that instance, if I do or , it will always fail to load the dynamic library. I haven't yet figured out how to launch the new instance and have the dynamic library properly loaded. So this must be something weird about my variables, some directory mismatch, or something else. I've already sent my here, but in case you haven't taken a look, I have in it (which in turn makes return ), in case there are some hard-coded paths that don't respect this nix setting and is messing up with the dynamic library lookup. Just bringing it up since there have been some Nix packages not respecting that setting and causing issues with me in the past.\nHey did you end up finding a solution to this? I'm experiencing the exact same thing, and this thread is the only place mentioning these issues\nUnfortunately, no, I haven't. Whenever I need to use MuseScore with MuseSounds (which is, all of the time), I launch MuseScore by doing in the terminal, which is one of the workarounds I've stated which have MuseSounds working.\nAlright, thanks! Yeah I've seen your comment above, and to my surprise did indeed work. Maybe I'll just overwrite the mscore.desktop file with this command for now (Btw, thanks for the debugging effort you've put tinto this)\nis not an issue due to the way library lookup works on Linux. Since is an optional dependency, it is loaded with (in via ), and says in particular that dependencies of such libraries are looked up in directories of the executable that calls . is a C++ application, so it also depends on and its contains a /nix/store directory with , and when it loads , the latter uses the same from the nix store. You can run to confirm that it has and that its contains the directory where it is located ()\nThe reason , etc. do not load is that the latter contains a shady function that prevents this library from loading when the calling process name () does not end with or . When you run MuseScore in the console from PATH, the process name is simply (without ), and the library declines to load. Similarly when you open a new score from MuseScore, it launches another instance of the program with the name which also does not end with .\nthe f. Amazing job figuring that out! (And also thanks for the MuseSounds nixpkgs PR )\nIndeed this is a very impressive investigation and conclusion . I'm a bit confused though - whether we use a wrapper or not shouldn't make a difference right? Because the wrapper script (whether it is or ) always set to right - hence the shady function will return , at least as a of right? I wonder whether we should perform a to an absolute path in to improve 's experience described here.\nI have already made a fix and will make a PR out of it very soon. MuseScore's argv0 will always end with regardless of the way it is launched, so can stay the same. (In my experience during system upgrades .desktop files are more convenient without absolute paths.)\nYou are right about the initial launch of MuseScore, but when you use File -New, it launches another copy of MuseScore, and it does this (as is common on Linux) by executing the target of , which is and this becomes of the new instance of MuseScore.\nThat's my experience as well. Ah wow, now I understand! Thanks. It means the patch will be much more involved... Good luck with that! I wonder why upstream decided to add that \"check argv[0]` in the first place. The 4.4.2 release just came out, and the build locally seem to go on OK, so I just opened - marked as a draft, waiting for your magic patch.\nThe API of is easy to learn from , which makes it technically simple to use MuseSampler without MuseScore. For example, someone writing a VST plugin could just load and use it as a virtual instrument player in any DAW, without MuseScore. I guess that Muse Sounds authors wanted it to be tied to MuseScore, so they made it fail to initialize when loaded into a program with a different name. This is particularly clever against VSTs since a VST can't change argv0 - the user of such a plugin would have to rename their DAW to !\nWow, thank you so much for the debugging efforts you did! I'm available to do some testing if you need any.\nThanks! You are welcome to test"}
{"_id":"q-en-nixpkgs-84935e7bc09895ea33fd1b5d390643b284094f24c934501080fb4f82be7260f6","text":"Project description for python! <!-- Describe the project a little: --It is built in reference with offical except it's color quantization part, which is based on thanks to . Metadata homepage URL: source URL: license: mit platforms: linux Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-dae4a8a1e068003d53d23d2c8ca4eadc80b4b0359c9fb674bf4e06410c7436e3","text":"When bulding an lxc template with and the , the ping binary does not work as the not root user, giving: Steps to reproduce the behavior: my config repo: the nix-secrets input (private repo). `nix build .#proxmox the resulting tarball and upload to proxmox as an lxc template a container based on the template in (without nix-secrets repo, password is 'nixos') and attempt to run ping. The ping binary successfully pings something. the Home-manager systemd service also fails, but that might be a separate bug, reported Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\non NixOS 22.11: On NixOS 24.05: Seems to have been broken by I'll spend some time over the weekend to see what the best way to solve this issue is."}
{"_id":"q-en-nixpkgs-f94b609e5fa01b4f98b72d60afa5e0be87707db30edb8b81f772c0772d0bdc97","text":"Package name: gruvbox-gtk-theme Latest released version: unstable-2024-06-12 <!-- Search your package here: --- Current version on the unstable channel: unstable-2023-05-28 Current version on the stable/release channel: unstable-2023-05-28 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-3e8775b4e786460c5506cd192a3d10dfaa170799d96f059320489176e00a5963","text":"Decoders fail to load due to . Steps to reproduce the behavior: Working Module Imports (no ) Working decoders Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nCan you check if the above PR fixes the issue. I'm not able to test it right now as I don't have access to a DSLogic device ATM.\nI'm unable to test the PR ATM, but the issue didn't require any device to be attached. Just install and run DSView as per bug report.\nOk just running and playing around did work for me after the update.\nHey, I checked this on my box, and I don't seem to be able to reproduce this issue, decoders seem to work for me when trying the demo device."}
{"_id":"q-en-nixpkgs-a160cab2464eb1639dfc46e64ad86c1dbde06d6765c2b90870d019a66ef24a41","text":"btrfs-assistant failed to use in nix-shell Steps to reproduce the behavior: A clear and concise description of what you expected to happen. nix-shell should success <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI fixed the nix-shell -p error but the application doesn't run properly. Looking into if we need to do some env variable nonsense because of the way the application works."}
{"_id":"q-en-nixpkgs-7bf75875c9d1b80f569b257b3159698028adf475f5c220428f0b4f10b487a3a7","text":"Steps to reproduce the behavior: Discovered via the Nixpkgs LLVM Workspace project. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --No maintainers, CC last person who touched Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-b88e092ccd37f1556336ab182bfad77631b3a16328d6020c527fdf8783a5338a","text":"Steps to reproduce the behavior: Add any other context about the problem here. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nMaybe we should just wait and see whether 2024.02 also has the same problem (that should be around the corner, see ) For now rolling back to to get unstuck.\nI'm skeptical about IDEA 2024.2 would use JRE 21, as in the following comment From https://intellij-"}
{"_id":"q-en-nixpkgs-83daa28c9c75e98bea8046f283ecd82d9f1f34e4b4e7dafee8d3385aaeffc87e","text":"Package name: Latest released version: 1.9.0 Current version on the unstable channel: 1.7.0 Current version on the stable/release channel: 1.2.0 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-c8993f7f8bc1e24f0121efb4baa786b6c50cf6bcf2ee87190ba595bda93968aa","text":"The package is broken, , because it fails during on Darwin. I am uncertain if this is something that has to be fixed in , or can it be worked around in nixpkgs? Steps to reproduce the behavior: Build on macOS <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\ncan you help out here?"}
{"_id":"q-en-nixpkgs-60786c6c1189cfb376d7a168b953be6866ac23746aeefdd1c2028e3a6924a0a4","text":"There is Bitwig1 and Bitwig2. I for example have only a license for Bitwig1. And I will not pay frequently for features I don't use (because I'm not a professional musican). Right now I use the overrideAttrs to install bitwig1, but it would be smarter to provide an older version as well (most likely there will be a bitwig3 someday). I could create a pull-request, just wanted to open a quick discussion first, because I'm new to NixOs and maybe there is already a very smart thing that makes live much simpler, which I'm not aware.\nSee . I think we shouldn't distribute paid updates for proprietary software under the same attribute without a way to opt out (other than rollback), this heavily breaks reverse compatibility. I think that should point to the latest version, with and pointing to specific major versions. Derivation for 2.x branch may override 1.x, since it has same dependencies and layout, the only difference is additional dependency on :"}
{"_id":"q-en-nixpkgs-5bd8e212e04c06193656033b93444a53018fb0ff77d0516ff3fed333b0ef34ef","text":"The package for the Ares emulator doesn't seem to include Vulkan packages so it will return if you try to run a Nintendo 64 game. Other cores that don't use Vulkan are unaffected Steps to reproduce the behavior: Ares N64 game error in console glitched or black screen The game works ! see similar issue: Ares issue before I realized it was likely NixOS: GPU: RX 6800 Driver install setup: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-5410e279a67092dfe25aab830322b293c49c61c802afa4ac27e62b75039187e4","text":"Package name: quickemu Latest released version: 4.9.5 <!-- Search your package here: --- Current version on the unstable channel: 4.9.4 Current version on the stable/release channel: 4.9.5 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --fedx- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-724e5becf91ad18990c1bc72fadf6618e9800ae61f2d816238b85ea704b967ae","text":"Specs-nvim does not work in Neovim 0.10 and the author has not updated the upstream version. The derivation should point to the updated fork instead: Steps to reproduce the behavior: specs-nvim with Neovim 0.10 specs-nvim errors Should I have used the out-of-date packaging report issue template instead? Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nno pb for me. Feel free to submit a PR. Pinging as he might know better.\nWhat does no pb mean? I'll make a PR\nI guess that it means \"go for it, let's switch to the fork\"."}
{"_id":"q-en-nixpkgs-d1617b105472423cdf9866b3ab79779c55c200a1ea392a8a728b9e2dd2616c19","text":"Package name: uutils-coreutils Latest released version: 0.0.27 <!-- Search your package here: --- Current version on the unstable channel: 0.0.25 Current version on the stable/release channel: 0.0.25 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-48bcc478953a3fec5e8cebf4d6f7f23a41c1686d043ef0f990e69f40bf251b4e","text":"Xpra supports forwarding invocations from the server to the client. The Xpra client defaults to hardcoding as the command to invoke on the client, which does not exist in NixOS. This bug does not impact hosts other than NixOS. Steps to reproduce the behavior: somehow. The Xpra client should invoke the local . N/A. This could be fixed by setting the open command to just , but it might match the intent better to change it to something like perhaps. Presumably, the point of doing it this way is to avoid any other copies of that might appear on the , e.g. Xpra's own. I suppose we can't really rely on to exist: even if we could on NixOS, we can't on other Linux distributions running Nix, where might make more sense. It seems to me that the best compromise, then, is to hardcode a path to copy of and users can override this using the option if needed. The open command that Xpra uses is configurable on the client-side, but the default is set as a platform-specific variable. On Linux, it is set in . Because there are separate values for both Windows and Darwin hosts, patching the value to shouldn't ever break them (but of course it should still be a Linux-specific patch, otherwise it unnecessarily brings into the closure.) Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-90d1d3f444922b7412199a0ad83b7755ccda3e5f2f963ce7b706bf89cf3d143d","text":"The src URL () for Azure Data Studio redirects to latest version (1.48.1), not the specified version (1.44.1), leading to build failure due to hash mismatch. i.e. redirects to . Steps to reproduce the behavior: Specified version of Azure Data Studio should be downloaded, verified and installed. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-c447e23af1d1e0688270b541c434f6411be4193190c6dbc6bc7e6c2403332857","text":"Project description goread is an RSS/Atom feed reader for your terminal. It allows you to categorize and follow feeds and read articles right in the terminal. Metadata homepage URL: source URL: license: gpl3 platforms: unix, linux, darwin Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-405a41be3778851b7f562b354dade4e77c6847562404b7366e50040317528a1c","text":"Klee installation returns the following error: Steps to reproduce the behavior: <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nThanks for the report; looking into it.\nLooks like a STP issue on master/unstable; any ideas?\nKLEE works on 24.05, try that in the meantime.\nSee also: , , and others. Seems like this is treewide.\nIt looks like there's an update to ; I'll bump it after gets merged but those two changes should fix klee."}
{"_id":"q-en-nixpkgs-9289b9589bc2f8ed7edd2f6f42bd24c849ed6e0fd2d968466fdb083e0bc3f689","text":"fails: see: cc: n/a <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\n(Fixed the title, my bad)\nThe current () tailscale builds fine for me on , but I gave it a try to be more \"correct\" anyway. I attempted to change to the following as a fix: However I seem to get the following error: I also tried running itself in the just to see what happens, and got an exec format error, presumably that's the cause of not having a completion in the first place."}
{"_id":"q-en-nixpkgs-0c23d7a6c22d37f49b4fb2eed1b7d983be23530835eec01f224aec5ea8cd69af","text":"Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-ab2c3a44aafb854d686d822521138ac9b1d4d3f3eb9b1c706bce5d5f37e5e073","text":"Flexget is not building on python3.12 on NixOS Unstable since nose 1.3.7 does not run on python 3.12 According to flexget website only pything versions 3.7 to 3.11 are supported: Steps to reproduce the behavior: to nixos unstable to install flex A clear and concise description of what you expected to happen. Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-fda2847a5efac115408102b7b6d296eb6110f13e89a2d0ae5cffef471325386d","text":"Package name: vkquake Latest released version: 1.22.3 Current version on the unstable channel: 1.22.3 Current version on the stable/release channel: 1.22.3 Notify maintainers Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-8f76b5679754758c372c85325c4d2a3fe696191e82e23bfe480974f7896e52e6","text":"Steps to reproduce the behavior: with One test is failing: only the relevant part Add any other context about the problem here. (note, I've set the project to the geospatial team, which should be enough to have it in the radar maybe?) Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI'm currently bisecting to confirm my intuition that it is a dependency version change that is triggering this failure.\nThe first bad commit could be any of: Probably an upstream issue?"}
{"_id":"q-en-nixpkgs-5ab01410c51416f1dccc6fc230a457cfb2e161a6e00850e61eb1ad77613294c1","text":"Rendering a project in Kdenlive with default rendering settings leads to the following error: Steps to reproduce the behavior: ( is the latest revision of at the moment of writing this) any clip to the timeline (for example, drag in an image) \"Render...\" \"Render to File\" You might need to use as your desktop manager to reproduce this issue. There is claiming that this crash doesn't happen on Plasma 6. I am using Plasma 5, but I am currently unable to verify if Plasma 6 would fix this issue (Plasma 6 is broken for me for unrelated reasons). Render should succeed. See issue for the previous instance of this problem. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nAre you using the kdenlive matching to your desktop? Edit: nope, you are using the plasma 6 package. You want to use plasma5Packages.kdenlive.\nis plasma6? Wow, that's quite a foot gun. Shouldn't it be called ? Also, fails with the same error.\nIt really should not matter, and will (once is entirely gone) be whatever versions upstream ships.\nit seems that the problem either wasn't fixed or has resurfaced again? I was able to reproduce the exact same error with and (where is the latest revision of at the moment of writing this)\nPlease try again after the staging-next merge (which should make its way to unstable in a few hours).\nI will. Could you explain why staging-next is relevant in this case, though? I am just curious. I checked using and in the github UI directly, and it seems that the changes from should be already present in revision .\nstaging-next reshuffles protobuf versions again.\nI wasn't sure which staging-next merge you were originally talking about, so I tried the same thing with commits (current nixos-unstable), (current staging-next) and (current master) and I am still getting the same error. Can you reopen this issue, or should I open a new one?\nHm. Actually, I think that I figured out, what the problem is. It seems that actually saves some paths in . For example, my had the following lines: Deleting the line fixes the issue, but then adds it back again with the new path, so it will almost certainly eventually break again. I am honestly not sure, why does do this, but we should probably either patch this behavior or report it upstream.\nThat sounds very wrong because you have defaults from two different kdenlive versions in there.\nI can tell you that I've never touched prior to this debugging session. I didn't even know it existed. Those lines seem to be automatically populated on demand. Most likely, these lines were simply generated at different times, and so they picked up different paths to the \"current\" version of . Also, the fact that these paths are different from each other is mostly irrelevant in this case. The real problem is that they are cached once and aren't updated when a different version of is used."}
{"_id":"q-en-nixpkgs-88dc43d79895e4dc8c3dbe7b4216f928dfa9408e9f996f138c794fa14118aec0","text":"Steps to reproduce the behavior: (or try building pkgsLLVM.llvmPackages_git.compiler-rt) It seems like useLLVM causes the sanitizers to be disabled, not sure why this is the case as LLVM can rebuild itself with sanitizers enabled (outside of nixpkgs). See Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nOk, removing those useLLVM checks resulted in this:\ncc may be helpful.\nUnable to reproduce this.\nHuh. I just ran into it with the new llvmPackages_git 19 -20 update in nixpkgs Hopefully the above nix file will be allow someone to reproduce it.\nThat seems to still produce the same result.\nTurns out this is related to .\nCC since this is related to sanitizers.\nIt's reproducible by building at version : Building version works fine:\nInteresting, I wasn't using no-libc I think but I was doing it under pkgsLLVM and on aarch64-linux. That version doesn't look right.\nReally? It seems to be the . Also, I have good news! Following the , I tried adding to , and it works:\nYeah, that's two weeks ago. Git is updated every week on Sunday. The PR for this week's update worked (although it only tests LLVM to clang). Interesting, I can get on this tonight."}
{"_id":"q-en-nixpkgs-86e8c265e7d7748db7f733043c71ce08fff65b3f5b3267a80f1054225e7f117e","text":"It would be nice if generated and options (only ). Processors without microcode update might run with misbehaving flashed microcode: /cc\nDo you mean like setting it to or just mentioning it? It's a good thing for sure to have working hardware as opposed to not working, but OTOH it pulls non free code...\nOf course setting it to true (actually to ), mentioning it would have no effect. First, already pulls in literal unfree code, e.g. Wi-Fi drivers on MacBook Air 6,x, not even redistributable. Second, non-free code is already running on Intel processors, I don't see any difference whether it's flashed or not (I know this is controversial). Third, for purposes of Nixpkgs, is free. And of course the kernel contains a whole lot of firmware blobs.\nStrongly disagree here as its an equivalent of setting = true; thus will taint every installation by default, which is a reverse policy of the current nixos from optin to optout. That said, being complete new user, I have myself figured and found these options myself. My system booted without them likewise fine. In general, AMD patches CPUs with AGEISA BIOS updates, so does Intel - however these updates are optional and up to the user. This is should be mentioned on a) Installation page b) NIxos wiki - not be made default option, because this action is in conflict with policy.\nI disagree that this is equivalent. CPU comes preflashed with microcode, BIOS comes with updated microcode for the processor to boot correctly (microcode that comes with CPU is very buggy and usually can't go through boot sequence correctly). It's not clear how it would add to the problem if we were to load an up-to-date microcode image. This is equivalent to shipping firmware to users that haven't set flag, which is current NixOS policy. Intel microcode is freely redistributable AFAIK.\nThe default state for these two switches is very clearly \"disabled\", plus they have no affect on non-AMD/intel CPUs. Flipping the switch in hardware-, is basically flipping the default state in the backstage, which is much worse than setting the defaults to \"enabled\", as it creates distrust and inconsistency cases. Further, firmware updates can cause real issues, not only solve them. It is irresponsible to think that they only solve issues. AMD AGEISA updates on 785G on has actually been broken for over a year for me, refusing to react to USB events on boot until I manually bypassed the tier 1&2 support (which is hell of a job for Linux user) to get to actual support for testing. About Intel, I have only encountered issues with gen3 PCIe cards in Intel motherboard for 2nd gen Core CPUs, but fixing this also required a year of waiting and UEFI reflash. Also speaking as an owner of HP() and Epson(sx525wd) printers - updating ucode actually caused hardware reflash and made legally purchased cartridges worthless. I am yet to encounter a CPU that can't pass boot sequence due to buggy ucode, but I doubt hot-loading it on OS stage would fix it, because it needs first to get to this stage. However I can clearly imagine the unpredictability if these switches are properly enabled.\nMicrocode is essentially a firmware for the CPU. NixOS, more specifically, the Linux kernel, will load firmwares for any device with support for that without asking the user. Loading microcode updates does not require a policy change, if anything, this would make the policy more consistent across the board. On a tangent, I don't understand why anyone would tolerate a blob flashed on CPU, and in BIOS, but would not tolerate the same blob being loaded by the kernel. That can be said about any software. Usually, the opposite is true. What I mean is that every BIOS except Libreboot contains microcode blobs that it loads onto the processor early in the boot sequence. Without that blob, boot sequence will most likely fail somewhere along the way. That's the reason why Libreboot doesn't support modern Intel/AMD CPUs, and why there is Coreboot/Libreboot split in the first place.\nYou are mixing very different cases into one bottle. Some hardware requires ucode to be loaded to function and won't function without. Some hardware has optional ucode updates. For this hardware, some classes load firmware update and then flash themselves - some lock the possibility to install previous revision, and some classes load the ucode as temporary override (which is case with AMD/Intel). Some hardware has open firmware (sources), but majority does not offer source. In essence, ucode is a blackbox and live-updating ucode is unpredictable black box. For some devices, ucode loading is a must, for others its optional. I suggest that we limit ourselves to specific case of Intel/AMD ucode, for which ucode updates are optional and crucial updates are applied only via hardware flashing. I am pretty sure that the reason why Libreboot supports less hardware has to do purely with acceptance criteria for how tainted (\"blackboxed\") the firmware is, and that both support little of x86, because Intel refused to provide specifications and AMD was open to some extent, but then closed up the AGESA after Phenom II because they most possibly implemented TPM in firmware(fTPM). About EULA-d + encrypted software, we are talking about closed-source hardware. For manufacturers of such software, surely opposite is true and the updates are usually used to keep their policy consistent and enforced, for users it means giving up and restricting all rights. That said, I found those options myself easily as a new user (yet my hardware functioned without issues without them) and I see a problem if they default to one value, but change \"automatically\" and unnoticed at run-time. Either way, this option should be documented better - and this includes places where its flipped.\nIt's not clear what difference does it make. Say, from freedom perspective, it would not be OK to use Windows even if I were to ignore any updates for it. If we ship updated microcode, which includes security fixes foremost, how would that taint installs any more than they currently are, with in-kernel blobs and such?\nIt has to do with blobs for the most part. See:\nIt replaces non-free code with a more up2date non-free code. Given Meltdown and Spectre this should be the default. Especially since it would be generated at installation time. Accompanied with the right comment, users who really want to have outdated CPUs will be notified and can delete one line. (well two with the comment).\nI would like this to default to (microcode essentially is firmware). isn't supposed to be edited, it should just use so that user can override it in their config.\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nstill think this should be done, as new users probably don't know this option exists at all and that you have to enable it by yourself in NixOS. With concerns on, this should at least be generated as commented out, with a hint \"while you will probably want this, note that it is unfree software\".\nI marked this as stale due to inactivity. &rarr;\nWhy is this no longer in discussion? As someone new to the NixOS universe and coming from Ubuntu I wasn't aware that I had to explicitely include this in my hardware configuration on NixOS until I stumbled across other people's configuration files. I definitely think it should at least be prominently featured in the installation manual or even better be enabled by default, as mentioned.\nOne way or another, needs to be enabled by default. These updates fix security issues, there are probably a lot of NixOS users now that don't even know that they're running insecure microcode. See also the discussion in\nI've never replied to the first comment! While I still support the open hardware movement (and thus free software firmware) I also believe that having QEMU misbehave because of a missing update is not really a good default option. If the user is concerned about the microcode updates or about managing security at this level then they will probably read the auto-generated configuration and immediately spot an and will comment it out :-) So thanks for the PR, I can't review because I don't know perl :-p"}
{"_id":"q-en-nixpkgs-d4e4db62dec1726fc9b3d4f7a7e7aeb3d754829e2bd0b96f38193dd5a70adde3","text":"Steps to reproduce the behavior: Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-160d7dce6e96f2fc12ba9927817e8f2994fb1168c89c0ec5d81b47bc69b59ca1","text":"Package name: nuclear Latest released version: 0.6.31 Current version on the unstable channel: 0.6.30 Current version on the stable/release channel: 0.6.30 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nMade a PR, would appreciate it if you tested it - I don't use nuclear very often at the moment."}
{"_id":"q-en-nixpkgs-db0c9ece9088934a11b6b4d828a1df55041d27d628559ef7ee4d2d6a44424c9d","text":"Project description Library to build personalized AI powered by what you've seen, said, or heard. Works with Ollama. Alternative to Open. Secure. You own your data. Rust. Has both CLI and desktop app. Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin, windows\nI am somewhat interested in this package and started a PR but I am super busy these days to thoroughly test the package. If anyone can do that, that would be awesome."}
{"_id":"q-en-nixpkgs-186d960f7b61df247b6b02b9ed58145152543375f1a14dadd108b84f45ef209b","text":"404 when trying to install. Version is out of date on both stable / unstable. Current version is 1.3.5 aka <"}
{"_id":"q-en-nixpkgs-59fae1a897df87821613174ca7ab9b6371f023642fe8f6db37fa50972707e9ca","text":"Steps to reproduce the behavior: Discovered by <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-e5e73279382255502e12085f7eae1e1ccd8acd423ee0a76e2e2c3e7e6bfbd301","text":"Ollama is a big project and I think having a nice UI in the stable repos would make nix even better. I can help with the PR but I am new to nix and don't really understand the git workflow and all Pinging the package maintainer: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nCurrent master is broken for (see ). However, I don’t think that should be a problem while backporting to , as .\nBackport PR for 0.2.4:\nTo track the progress of backporting to latest , see:\nI love this, thanks!!"}
{"_id":"q-en-nixpkgs-2fbe3fa36e9f47b32209ed2b1dfd9eb263bb4909681a349238f85f7de6af92ad","text":"Package . This seems to be interesting, we should package it. This issue is a friendly reminder for myself to package this some time. I do not feel offended if someone wants to do this, though!\noh wow, thanks!\nOh! I thought was the one who filed this request haha, my mistake! That's why I included as maintainer originally.... I thought there was a sudden change of heart! :laughing: Anyway, no problem GLHF. (and it looks like 1.0.2 is out now, if there's a compelling reason to update LMK otherwise I was going to see if they stabilized a bit before updating)"}
{"_id":"q-en-nixpkgs-a557822469ea8200c64e828266cfd5f879d743f783064bcea1681d73c85ba1b1","text":"After merging today, no longer builds. Not sure how to fix, :(. Building against parent commit of merge (not from staging), , produces working pandoc. Hydra log:\nLooks like changing to : But I'm not at all sure that's the correct sort of dependency."}
{"_id":"q-en-nixpkgs-1db9097f46fb331477f74d5e7accfb8155dd577ab5e07ab8f341ba64dde03037","text":"Hi there! I’m sorry to bother you with a ping. We realized that in the form of . All of our Nix expressions are meant to be MIT‐licensed, so this is a big problem, especially as packages using / may count as derivative works, potentially jeopardizing the licence status of Nixpkgs as a whole. We would therefore like to get permission from all contributors to this code to relicense their changes under the MIT license used for Nixpkgs. If you have any questions or concerns about this, please ask them in so that this issue can purely be for tracking the sign‐offs we’ve received. To agree to the relicensing, please reply to this issue with the following verbatim text: [x] [x] [x] [x] [x] [x] (confirmed via email) [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [x] [ ] [x] [x] [ ] (GitHub account is gone, although as this was just a treewide removal with no other changes it’s unlikely to be copyrightable) [x] Authors of unmerged pull requests in the old repository (just in case we ever take code from them): [x] [x] [ ] [x] [ ] [x] [x] [x] (I have not checked for unmerged PRs to this part of Nixpkgs yet because it’s a pain.)\nApparently I can only ping so many people at once; please look at the original issue text, sorry!\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs. PS: nice catch and thanks for taking care of it\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs. Like, sure, we still use it in our toolchain, and this move is very nice.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , and the pkgs/development/tools/yarn2nix-moretea directory of the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs or any other OSI considered free software license. IIRC if we get everyone else on the list, then will be fine.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nJust another gentle ping; if you could check the original post at respond as requested it’d help us resolve this licensing issue.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nWhile it wouldn't hurt to get approval from the last 4 people, I don't think we need that to move ahead: gebner only authored , a minimal change that I don't think would be copyrightable egasismus only authored , another minimal change that I don't think would be copyrightable chawki008 only authored a PR that was never merged: eqyiel also only authored a PR that was never merged:\nYeah, I agree those changes aren’t copyrightable. I did email the authors of them but I’ll proceed with opening a PR to do the relicensing now.\nI agree to license all of my past contributions to , , and the pkgs/development/tools/yarn2nix-moretea directory of under the MIT license used by Nixpkgs.\nThanks so much to everyone for helping get this done so quickly, and for your contributions to the Nixpkgs ecosystem"}
{"_id":"q-en-nixpkgs-96c31bb12b81a6d4a1c084a6cddd0334daa5cd851c1f93db3bec84b15f5e76f0","text":"Hello, please find below a bug report about the Nix package qdidgidoc: I can not sign any more since the 24.5 update. Failed to sign container. Please check the access to signing services and network settings. Failed to sign container. Failed to connect to host: 'dd-' DECODER routines:0 error:1E08010C:DECODER routines::unsupported BIO routines:0 error::BIO routines::system lib I also checked if I am able to authenticate with https://web- . Both, authentication and singing works. For completeness, here my diagnostics output: Locale: English / enUS.UTF-8 Application version: 4.5.1.0 (64 bit) OS: NixOS 24.05 (Uakari) (x8664/x8664) CPU: AMD Ryzen 5 PRO 4650U with Radeon Graphics Kernel: Linux 6.6.44 -NixOS SMP PREEMPTDYNAMIC Sat Aug 3 06:54:42 UTC 2024 x8664 Libraries: QT (5.15.14) OpenSSL build (OpenSSL 3.0.14 4 Jun 2024) OpenSSL current (OpenSSL 3.0.14 4 Jun 2024) Arguments: qdigidoc4 Library paths: /run/wrappers/lib/qt- 15.14/plugins; 15.14/plugins;/var/lib/flatpak/exports/lib/qt- 15.14/plugins; 15.14/plugins;/nix/profile/lib/qt- 15.14/plugins; 15.14/plugins;/etc/profiles/per-user/getreu/lib/qt- 15.14/plugins;/nix/var/nix/profiles/default/lib/qt- 15.14/plugins;/run/current-system/sw/lib/qt- 15.14/plugins;/nix/store/h8m7pcd6d20jsx7hn81rsgkq0h884893-qtwayland- 15.14-bin/lib/qt- 15.14/plugins;/nix/store/vc9wklk5w3cskkymp52g7js83my4d9ch-qttools- 15.14-bin/lib/qt- 15.14/plugins;/nix/store/plix80gx31963j2qpsws7rid1gfr6cyn- qtdeclarative-5.15.14-bin/lib/qt- 15.14/plugins;/nix/store/h32d5q8ngcy68nddbw5vja3r316s20p3-qtsvg- 15.14-bin/lib/qt- 15.14/plugins;/nix/store/fbddznz5ln765jbl2i2mfnf8h7rkkmd9-qtbase- 15.14-bin/lib/qt- 15.14/plugins;/nix/store/h4li58h9pni11wigz8xldad9kgzc0q3c-qdigidoc- 1/bin URLs: CONFIGURL: SID-PROXY-URL: https://dd- SIDV2-PROXY-URL: https://dd- SID-SK-URL: https://dd- SIDV2-SK-URL: https://dd- MID-PROXY-URL: https://dd- MID-SK-URL: https://dd- RPUUID: is set by default TSLURL: TSAURL: http://dd- SIVAURL: CDOC2: CDOC2-DEFAULT: false CDOC2-USE-KEYSERVER: true CDOC2-DEFAULT-KEYSERVER: ria-test TSL signing certs: Patrick Kremer (Signature) European Commission European Commission JEROEN ARNOLD L RATHE APOSTOLOS APLADAS CONSTANTIN-ADRIAN CROITORU TSL cache: (65) eu-lotl-pivot- (335) eu-lotl-pivot- (341) eu- (343) g0xdp6w34ric1mdh8g7r0v8h85idkcg1-eu-lotl-pivot- (300) Central Configuration: Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nI can reproduce. The \"Signing\" animation is shown for a while, and then an error too: The error message mentions not being reachable: ! This hostname does not even have a DNS record. I spotted changing the \"Default TSA URL\". Maybe they updated the URL in some places, but forgot updating elsewhere? I found this URL in my qdigidoc settings too: ! Changing it to the new URL unfortunately also didn't help, as it doesn't like the Digicert certificate. Maybe unrelated, maybe not - I also see a in my logs. This looks like an issue in the application itself / some config shipped somewhere, but nothing NixOS-specific. I propose opening an issue upstream in qdigidoc, maybe linking to here for context.\nI tried manually bumping , and I tried using from . It mentions a signature to be invalid: Opened an issue upstream:\nI got a link to , mentioning a workaround. However it looks like the libdigidocpp changes linked there are not a sufficient fix, so I cannot patch this in nixpkgs, and it for now requires each user to manually set that address and import certificates.\nAs I understand now fixes this issue with the related commit: . Do you consider a patch release?\nUnless I'm missing something, is the only relevant change for us from this commit (which sets the flag). The patch doesn't apply, but IIRC I tried setting this manually via and it didn't fix things for me back then. I can try again tomorrow or so, re-rolling that patch and trying to sign again with a fresh profile. Yes, absolutely, if there's a patch we can apply to fix this for nixpkgs users, at least when starting with empty state I'd like to do it.\nI did apply the TSA_URL change in , and it seems this was enough to get signing to work. PR is up at , please test and report back there."}
{"_id":"q-en-nixpkgs-773afaab7c50db48203f5f1788c757aec8114a90f6265444443249c37d31e183","text":"Project description A modern top command that charts system metrics like CPU load, network IO, etc in the terminal. Metadata homepage URL: n/a source URL: license: mit platforms: darwin (amd64 and arm64), untested on other platforms Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-3f8326be7db97bd2284f8f68436a6dd36b61ec5237a3c08cb1888e3618450cf8","text":"Not too long ago was updated to have as its dependency. However both python packages have a , so the created from can't successfully build. Steps to reproduce the behavior: to build error: Not exactly sure what should be done, since I'm not familiar with the cli tools. As a workaround you can probably do (listed in meta.maintainers) (author of the that the dependency) <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-5e1aeaff6a38e8428d47a64fd620867b91805e3ffd34c71f9d06ab387bf5859b","text":"Package name: tailscale Latest released version: 1.72.0 <!-- Search your package here: --- Current version on the unstable channel: 1.70.0 Current version on the stable/release channel: 1.66.4 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-8647570b6161614f3304f8cdc85cfb063f2c5a84685720d56b7c73e29c3ddf9b","text":"Package name: gpt4all Latest released version: 3.2.1 <!-- Search your package here: --- Current version on the unstable channel: 2.8.0 Current version on the stable/release channel: 2.7.5 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nHi, I don't maintain GPT4All any longer, I switched to some other technologies (Open-WebUI).\nhaha, maybe i'll give that one a try then :) ty for all your hard work"}
{"_id":"q-en-nixpkgs-f5f56d61f9c4514bc1cdffb9efdb068834ccc1d8cea571edae3f22a823689835","text":"Steps to reproduce the behavior: X Add any other context about the problem here. <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nsee"}
{"_id":"q-en-nixpkgs-2c9d61fba201bcf378ec7264fd92ccf137efc4566ba24da675ac66fb1d3d587f","text":"Getting an error trying to build grpcio 1.7.3 on darwin make: Circular /private/var/folders/4k/jpntp1hn6b7c9ymh7v044b9h0000gn/T/nix-build-python2.7-grpcio- <- /private/var/folders/4k/jpntp1hn6b7c9ymh7v044b9h0000gn/T/nix-build-python2.7-grpcio- dependency dropped. ar: illegal option -- / usage: ar -d [-TLsv] archive file ... ar -m [-TLsv] archive file ... ar -m [-abiTLsv] position archive file ... ar -p [-TLsv] archive [file ...] ar -q [-cTLsv] archive file ... ar -r [-cuTLsv] archive file ... ar -r [-abciuTLsv] position archive file ... ar -t [-TLsv] archive [file ...] ar -x [-ouTLsv] archive [file ...] make: * [Makefile:8459: /private/var/folders/4k/jpntp1hn6b7c9ymh7v044b9h0000gn/T/nix-build-python2.7-grpcio-] Error 1 nix-shell system: host os: multi-user?: sandbox: version: channels(joestein): nixpkgs:\nTry:\nThanks that worked to get past the AR error, fighting another error now with libtool but progressing along again. Thanks!!!\nThis derivation works now"}
{"_id":"q-en-nixpkgs-5bd542d421fc9bcf34a7c0a21774c613eacd982df96aac1d8db510b54ee79a90","text":"Package name: rio Latest released version: 0.1.10 <!-- Search your package here: --- Current version on the unstable channel: 0.1.7 Current version on the stable/release channel: 0.0.39 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. -- Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nwill do a PR this evening"}
{"_id":"q-en-nixpkgs-91794dcd6bbf162b393294d26d6e1bfe6d7d25a492908e593a796c64aa60c2e2","text":"Package name: ffmpeg-normalize Latest released version: 1.28.3 <!-- Search your package here: --- Current version on the unstable channel: 1.26.1 Current version on the stable/release channel: 1.26.1 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers <!-- If the result shows no maintainers, tag the person that last updated the package. --Note for maintainers: Please tag this issue in your PR. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-18fa2dc6bdd94b19fb737e87d3c54311b6a76df29ad51256aa104462a559da83","text":"Steps to reproduce the behavior: dvc another dvc hash error, i dont know why <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-995558cf18e491320129dce4bf7037ef835175b233c5e39c5f19900eaea690db","text":"First failing build (4.9.76, but not always): Latest (4.9.77): Looking through the changes for any obvious culprits...\nbisected to (as the first \"bad commit\"). cc\nReverting the commit on master fixes the build, but need to replace with to at least fix evaluation on cross. Since doesn't seem to build on cross (at least not on any architectures I've tried) I can't tell if this \"works\", but would be good to fix this on non-cross regardless.\nSee corresponding PR, this has been handled. Thanks!"}
{"_id":"q-en-nixpkgs-8571a793b3fffc56298e670d9ef77049a92fcd3de22793fb38116ce6e4e01aa1","text":"Project description <!-- Describe the project a little: --Syntax Tree Surfer is a plugin for Neovim that helps you surf through your document and move elements around using the nvim-treesitter API. Metadata homepage URL: source URL: license: mit platforms: unix, linux, darwin, ... Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-9094011f0db5c3c7fa214d5fd52c142e265fa1a795639f29e6ebc3e29b7e4ee2","text":"is currently using this command creates the temporary directory inside the current working directory instead of in . This can cause the generated directory to be included in the final package, as I'm assuming accidentally picked it up. I encountered this while working on and The temporary directory is not included in the output use use (output is of the format ) note: this isn't really used across nixpkgs use something else <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-c7ce5aaa0f5a869f91edc6c2f0388768ca62bbb2ac83684bc34f9ad7285b3199","text":"The program described by this man page is not available in master. ag ipvsadm\nPoked at this, looks like it needs which because it was old and krufty even back then (note the patches required). That seems like a bit of a show-stopper, unfortunately.\nThanks! Is there a replacement for that I am missing? is still a part of the Linux kernel.\nEnjoy! LMK if you're able to actual test it or know how to employ the tools :+1:.\nI plan to do that tomorrow. I hope the tools will tell me the crucial information I need to make it work.\nThis works great and should be merged. It told me something unexpected, but that's good."}
{"_id":"q-en-nixpkgs-3ebcd12adc327710f6c5f77937482997ffb7f47dfcf8f9ec5e5984b960042574","text":"Package name: libssh Latest released version: 0.11.1 <!-- Search your package here: --- Current version on the unstable channel: 0.10.6 Current version on the stable/release channel: 0.10.6 <!-- Type the name of your package and try to find an open pull request for the package If you find an open pull request, you can review it! There's a high chance that you'll have the new version right away while helping the community! --- Notify maintainers Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:"}
{"_id":"q-en-nixpkgs-70243f87d588be9c63ee14313352508055fdfe2350233ad6a8957a98812eb2a2","text":"Steps to reproduce the behavior: -rebuild test on settings.datadir = \"/mnt/garage0\"; resolves fine Tested on nixos24.05 but code of the module is unchanged in nixos_unstable. I'm using flakes as well Please run and paste the result.\nLooks like doesn't apply, maybe should tinker a bit on my own\nIt looks like the option type was adjusted, but not the other code in the module that checks that value. Shouldn't be too hard to fix."}
{"_id":"q-en-nixpkgs-d5e9156f156bd91ebc6fcfdb185e53da8f1901af07ed0d3d0afcb82c94fd6f7a","text":"Steps to reproduce the behavior: try to install/build the package using nix-shell or nixos the Domain is no longer configured. Teamspeak Download Pages redirects to this place: Please run and paste the result. Add a :+1: [reaction] to [issues you find important]. [reaction]: [issues you find important]:\nYou might want to check out\nNot merged yet\nAs a workaround, until this is fixed in nixpkgs: The file is still available at the Internet Archive. You can fetch it from there and write it into the store, with something like this:"}
{"_id":"q-en-nixpkgs-80acfa9bea337068279bed4d1f9279d2f42b5b67977871c04696fbfb5ca5e5a6","text":"cc system: host os: multi-user?: sandbox: version: channels(davidak): channels(root): nixpkgs:\nThanks!"}
{"_id":"q-en-nixpkgs-146da02b6d8147b961d73ae2270ce223ba659c6a1f1953ea75b2a08db1020be2","text":"The recent changes seem to make sage crash when it tries to run some code using openblas (I think). I installed the sage derivation from the latest github repository. In the sage interpreter, run the following sage crashes with the following message: Please run and paste the results. system: host os: multi-user?: sandbox: version: channels(okideal): channels(root): nixpkgs:\nProcedural: as it is , please specify the CPU model. Looks like the Sage build system had the idea to use somewhere… I.e. the local build could work. I wonder how feasible it is to find where Sage does it — the Sage derivation is allowed to rebuild a lot of dependencies because it is just too much pain to make it use the system versions.\nHere is the output of lscpu\nFrom the traceback it presumably crashes here: which is which indeed is an AMD-only instruction (MMX? thought that would be long obsolete by now...)\nYou were right. A local build works fine. Thanks for the help.\nThanks for your report. do we set to and give up?\nThis should be fixed by exporting (see ). I'm not quite ready to give up (even though it would probably be the healthier choice :grin:). I'm actually currently experimenting with doing it the \"right way\" and replacing the sage buildsytem with nix (with a mix of sage and system dependencies).\ncan you confirm that the latest binary works?\nSorry for the late answer. The binaries from the master branch seem to work. The initial crashes are gone and running gives only unrelated failures. Installing from still gives me the old, faulty binaries though.\nAre you sure you updated the channel? What do you mean by \"binaries from master\"?\nI meant that installing from the master branch of the git repository works. I was not aware that there is a difference between nixos-unstable and nixpkgs-unstable. The latter does not seem to include the fix yet. Installing from nixos-unstable works well. Thanks for your help.\nOkay, I guess hydra is a bit overwhelmed by all the fixes for the upcoming release. Thanks for reporting."}
{"_id":"q-en-nixpkgs-173a81a1fed75dc5a596c8a50c9d34093c02ebed26a348d19a9199e0682c72ec","text":"When I trying to install twinkle, I recieve such error:\nI tried to fix twinkle a few weeks ago, and I did get it further, but I didn't fix it completely. I can try to find the WIP this evening if someone is interested.\nI managed to fix it in the end. Details are in the commit message."}
{"_id":"q-en-nixpkgs-b367a9400dee57292a1c42c414e77d6d3ea10269a2cd2ff3d10b1fdfe1db795d","text":"Cockpit is a server manager that makes it easy to administer your GNU/Linux servers via a web browser. Website: http://cockpit- Upstream: License: LGPL2.1 Arch Linux package: !\nCould be a lot of effort to port: ... I have not looked at the code how many assumption it makes about the system, but in terms of system configuration nixos is a special snow flake.\nProbably only a subset will work. But I think there is some stuff in cockpit that is helpful even on NixOS: View journald logs and systemd services Manage virtual machines via libvirt Get a shell terminal from the Web UI Some general system information like disk usage etc. It's mostly read only stuff but I'm not sure how easy it is to only use parts of cockpit that would work.\n+1, would love to have this\nGiven that NixOS is pitched as a server os first, this would be very helpful. I already tried to but got stuck because it requires a few npm packages\ntry with\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . 3. Ask on the on .\nI still think it would be good to have this in NixOS.\nWell as pointed out, there will be quite a lot of patching up paths to do, as well as the NPM packages that need installing (try yarn2nix or npm2nix for that) Seeing as it's non-trivial, it will have to be packaged by someone with real interest in getting it, and we can assist (make a draft PR and tag us on questions)\nnode is only required for the dev environment I have made packaging cockpit, it still fails to compile tho Any help would be greatly appreciated\nbuilding build flags: SHELL=/nix/store/5hplxs3kprdp6v62qk3di6a1m1s1paiq-bash-4.4-p23/bin/bash GEN src/ws/mock-dbus-tests.h Traceback (most recent call last): File \"/nix/store/5zajbm8674d93nix4h8b0hxgglj2pbir-glib-2.64.3-dev/bin/gdbus-codegen\", line 53, in <modulefrom codegen import codegenmain File \"/nix/store/5zajbm8674d93nix4h8b0hxgglj2pbir-glib-2.64.3-dev/share/glib-\", line 24, in <moduleimport argparse File \"/nix/store/jli50q24fvik6pfsnjs1jd90g3dy4p6y-python3-\", line 88, in <moduleimport re as _re File \"/nix/store/jli50q24fvik6pfsnjs1jd90g3dy4p6y-python3-\", line 145, in <moduleclass RegexFlag(enum.IntFlag): AttributeError: module 'enum' has no attribute 'IntFlag' make: * [Makefile:: src/ws/mock-dbus-tests.h] Error 1 error: --- Error\n----------- nix-build error: --- Error --- nix-daemon builder for '/nix/store/mdiihwskfp862y17wp70skdy74kh33v5-cockpit-' failed with exit code 2; last 10 log lines: File \"/nix/store/5zajbm8674d93nix4h8b0hxgglj2pbir-glib-2.64.3-dev/bin/gdbus-codegen\", line 53, in <modulefrom codegen import codegenmain File \"/nix/store/5zajbm8674d93nix4h8b0hxgglj2pbir-glib-2.64.3-dev/share/glib-\", line 24, in <moduleimport argparse File \"/nix/store/jli50q24fvik6pfsnjs1jd90g3dy4p6y-python3-\", line 88, in <moduleimport re as _re File \"/nix/store/jli50q24fvik6pfsnjs1jd90g3dy4p6y-python3-\", line 145, in <moduleclass RegexFlag(enum.IntFlag): AttributeError: module 'enum' has no attribute 'IntFlag' make: * [Makefile:: src/ws/mock-dbus-tests.h] Error 1 Seems like an error in the package itself.\nThe enum thing is due to a stray enum34 dependency in - enum34 is a backport of the enum module, and breaks things if it's installed on Python = 3.4. You'll need to look at why pyscss is pulling in enum34 - it's not immediately obvious to me from pyscss's , it looks correct. A pthread link error happens if that's fixed or if Python is removed from the deps entirely. That pthread link error is, I believe, a problem with the Cockpit makefiles; I made to fix it. With that change applied, I get I don't know anything about the Javascript packaging, so no idea how to progress from here. Good luck!\nIt compiles but gets stuck writing it's systemd.service files. I don't have much experience packaging applications which hook into systemd\nHow’s the effort? :crossedfingers: :crossedfingers: :crossed_fingers:\nIf this gonna work I could start replacing my with\nI have updated nix to solve the issue of beside updating the version and it built successfully\nI built the derivation posted by by including in my . It built successfully, producing . The output includes many things for like systemd units that do not get installed on the system - I'm wondering if these should be symlinked in as-is, or do they need to be converted to Nix? Most systemd units look OK but will need a fix to replace a reference to :\nIf is changed to (in patch below), and is to , the units get installed but are not enabled. The package also has references to and , please let me know if you know how best to patch these: I patched the reference to , and replaced with just due to build warning. Here's what I have, building entirely on what others have contributed to this thread:\nI considered running the docker container, but it requires and proceeds to modify the host's PAM config. Not something I personally like to see in a docker container... Back to the Nixpkgs. I was able to use so that has references to coreutils programs and patched the certificate paths to be . was because of patches to . Right now it fails running with SIGSEGV. I thought it might be systemd's setting, but it does the same outside invocation by systemd. I tried building newer versions of cockpit, but I kept running into There are some notes that might be helpful. Here's the latest packaging attempt: Even though I've posted like 4 times in a row, I think I'm done trying to do this. I hope someone else can pick up the effort.\nI was able to get past this by changing the interpreter in from to . Note that, for some reason, doesn't work. You also have to run , because Nix resets its date and that breaks the \"is packages- newer than \" check. I also had to replace with in . What's odd is that all of the scripts run correctly when run from . When this is done, will just error with \"Cannot find module stdio\".\nSpent some time learning nix by giving this a shot and actually ended up getting things to start. A lot of things are not working, but some might be possible to fix. For anyone interested I have left things here:\nAnother take on it Tested features: Logs Virtual Machines (cockpit-machines module, only managing works, creation of new VMs doesn't work) Services Terminal Code:\nTried to create a virtual machine, no luck, OS list is empty. Cannot create VM without choosing the operating system. I tried to fix it but I got the following error: !\nI came across this discussion in the upstream project and figured it might be helpful to cross-reference:\nYou require libosinfo, and osinfo-db\nI would to keep working on this issue, any hints you got for me would be welcomed. I do see any errors in my log\nI am also working on this Here is what I have so far: It's building only. I was able to solve some documentation build bugs.\n:eyes: !"}
{"_id":"q-en-nixpkgs-c4704cd635ae652281836f07d191ba9c282f742764fa6d1f67f1beee5af76a06","text":"Currently, the retpoline mitigation is only available in gcc7 . However, it still exists some old kernels that only compile with , such as those in Ubuntu Trusty LTS (kernel 3.13). Moreover, kernel modules loaded on these kernel have also to be compiled with the retpoline mitigation. Ubuntu (and Debian) integrated retpoline mitigation in their releases . Latest Ubuntu 3.13 kernel now requires kernel modules compiled with retpoline mitigation. They can no longer be built with Nix gcc toolchain. What do you think about backporting these patches? Someone could help on this?\nsounds good to me."}
{"_id":"q-en-nixpkgs-4b94648bb9b6436ce6bb25640e846f98a86ec22edf19918ea66a3cfe9ec5524f","text":"When saving a plot figure to pdf in Octave, the program panics and aborts Reproducible with both and from . If the GUI is used, the program exits with signal 6.\nDoes suffer from the same problem?\nYes, same problem with .\nRequires in , which we have not packaged yet.\nthis should be reopened, octave lacks in so the issue still stands. I think should be included in both and . Mainly so command line users can still save graphs to a file. I can make the necessary changes eventually or if someone else wants to that would be nice.\nOops sorry"}
{"_id":"q-en-nixpkgs-555094509a12335e1ce16beade9a4912ac08af9ba5a91a10dd3af9c1121e701b","text":"The MTU of the ethernet should be set to 1460 on google cloud platform, but it is set to 1500. For me this has occasionally, but not always caused the ssh terminal to hang when lots of data is being printed, and I assume could be causing other people issues. from - Create a google compute engine nixos image from observe the mtu on eth0 is 1500. Set the following: to fix the issue. system: host os: multi-user?: sandbox: version: nixpkgs:\nI created based on what you wrote but didn't test it.\nThanks, that fix was enough for me, though I don't know if it works in all situations, or if DHCP is not picking up some parameter like it is supposed to. I don't really know how to proceed with a fix for sure, but I thought I would document the issue."}
{"_id":"q-en-nixpkgs-9515f3f8f6afc0b3b94497e7251882fe0f6d60f84ada4c95fcdfa0ab1c7e47d5","text":"Darling is a compatibility layer for running macOS apps on Linux (think Wine for macOS). I think it would be interesting to work on with macOS support. They have reverse engineered some headers that Apple doesn't release publicly so having them in \"libdarling\" could be useful.\nWe already have this part:\nJust heard of Darling. I would be delighted if it could be used for running builds on NixOS!\nIt seem to be packaged by but only for macOS. If i understand the project correctly, that makes no sense. We want it on linux, to run macOS apps. Also, the package has no version. That's probably because they havn't tagged a version yet. So this is pre-alpha state software. could you update the package to the commit the \"stable\" AUR package uses and add a version as described here: ? Then we can close this issue.\nIt looks like the project has started using tagged versions:\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . 3. Ask on the on .\nI'd still be very interested in seeing this , especially with the hope of future cross-compilation support\nThis would even make it possible to package Mac binaries for Linux!\nI'm interested in this so that I can test cross compiled binaries when running tests.\nI marked this as stale due to inactivity. &rarr;\nNot stale!!\ndarling-dmg is just part of . To be exact, it is included as a submodule in the source tree of darling. The darling build will come with kernel modules. It would need someone with kernel-packaging experience to pack it into Nixpkgs. Build instructions in the Darling Docs:\nI marked this as stale due to inactivity. &rarr;\nI'd still like to see this, though largely for much the same reason Andrew mentions at this point -- I've started working with zig and being able to test cross-compiled macos binaries would be brilliant.\nI have got most of a Darling build file running, but I have found an interesting issue. Wever I build I get the following error: From this cmake file: As far as I can tell, the command is a special MacOS thing, and I unfortunately do not understand CMake well enough to fix it...\nWell, turns out the issue is that Darling relies on git-submodules, and it looks like this must be built against the Git repo directly, not against the released tarball...\nWhat would therefore be a derivation?\nWhat do you mean It's just a with .\nI tried to re-package the prebuilt deb from darling GitHub releases. Here is the nix code darling- Save these 2 snippets as and darling-, then run . With 2 problems: needs setuid/setgid perms, while nix-build would fail if dpkg-deb -x executed directly. I use fakeroot to ignore this problem. always fails. I am not a darling expert, have no idea.\nAh, fantastic, thanks! Looking forward to having this in nixpkgs anyway, but let's see how we go there.\nIt could be named when entering Nixpkgs.\nGood suggestion. I've edited those two scripts.\nI got as far as finding out that darlingserver doesn't start with that script because it expects /usr/bin, so updating my to a version, the next issue was that darlingserver was trying to mount overlayfs on \"lowerdir=/usr/libexec/darling,up\" (may be incomplete, this is strace output) which did not exist.\nAfter a bit of tinkering I got Darling to build from source. I opened with more info. Please give it a try!\nIt seems that they now use instead of the kernel module. This would make packaging easier.\nWhat's the purpose of ? If it is only the sandbox part of Darling, a new name like would bring less confusion. Update: Opened PR to do the renaming."}
{"_id":"q-en-nixpkgs-e79d486ed499c6348e660194e706a411f7686c5ffc3ab64b75c3aee06cf7785f","text":"The bullet points under Chapter 2, section 8 (the fdisk process) don't match up with what the real partitions are after write. The last four lines of fdisk setup are relevant: Earlier in the fdisk process, we create the partitions with the following numbering: I believe the re-numbers the partitions based on their location within the partition label. Since the boot partition is at the front of the disk and our partition is at the end, fdisk reorders the partitions as follows: The problem is when making the file systems. The manual still uses the initial partition numbers instead of the new, ordered numbers. Thus: Creates an ext4 partition that is 512M big and is supposed to be boot. And: Creates a FAT32 partiton that's supposed to be our root labeled . Swap isn't affected being in the middle.\nHi! Closing this issue since the partitioning instructions are now re-written for 18.09. (See ) As far as the ordering of partitions is concerned, the newly re-written partitioning instructions were written to ensure it wouldn't be an issue."}
{"_id":"q-en-nixpkgs-1479f85f325b24c7aaad0bdb50aee19e8e7295e53f9c3731c4bf14ffe3118243","text":"has been broken on master for a while now, blocking orca. I opened an but they are not sure how to fix it.\nI removed it from orca in"}
{"_id":"q-en-nixpkgs-e2b0ffa4d4f16e0a18332b0459e74d097ca74b762d50a02e42ff02c2d3929a25","text":"Tried to make a system upgrade, got this: Obviously, upgrade can't be completed. have in . run Please run and paste the results. system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nThis looks like the offending line: This could be related to what file system you are using?\nActually this looks like unrelated to glslang. Works on master:"}
{"_id":"q-en-nixpkgs-e0948e98378b7d21827ee94f0d2fdb392c0c6eec5a44967d44afccbae2aae32d","text":"does not build after \"boto{3,core}: {1.6.0,1.9.3} -{1.7.9,1.10.9}\" by system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nWhoops, sorry about that! I keep forgetting they're all linked"}
{"_id":"q-en-nixpkgs-45769e7d4185a4221862dac9b78843cb40bbab975881709dddf34e6b72fc2beb","text":"Open a gnome-terminal and go to any directory other than $HOME. Right-click in the terminal and click \"New Terminal\" A new terminal opens, but the current directory is $HOME. This has probably always been like this in NixOS. On traditionally systems is sourced which exposes the shell $PWD to the terminal. In NixOS that file lies \"dormant\" in the store: Users can workaround this by manually sourcing the file: (probably best to pick the one in gnome-terminal's clousure) system: host os: multi-user?: sandbox: version: channels(ole): channels(root): nixpkgs: Gnome installed by\nIs that supposed to be done for new windows? It works for me with new tabs, I think.\nIt does after sourcing the file so I assume it's intended. I'll check on an other system when I get the chance. I get the same behavior with new tabs (does not preserve $PWD when is not sourced) Can you check the content of (bash) or (zsh)?\nI use fish-shell and the pwd just seems to be preserved.\nSeems fish notifies various terminals . More context:\nWould something like the following work? This global solution is not very nice but I am not sure how to address it correctly.\nThe alternative might be to add something like the linked fish config to : The script checks if the shell isn't bash or zsh, so it should be safe to do so. Did a quick test in a VM and it seems to do the trick.\nFor some reason the above snippet doesn't \"initialize\" for me. I have to manually run . Weird. I even some manual trace and see that is sourced. But still it doesn't work. I'm using bash, but perhaps I have some untracked stuff in .bashrc that is interfering? I'm currently unable to debug using the NixOS VM. stops in stage-1 at \"mke2fs 1.44.1 (24-Mar-2018)\". EDIT: typo\nhmm, weird, I'm specifically running this diff on unstable: I'm able to launch bash (from zsh) and the directory successfully updates when creating new terminals, though I can't quite remember if I tested things out with bash in a VM. edit: managed to do ctrl-return\nThanks to I am now able to use \"nixos-rebuild build-vm\" again (on NixOS 18.03). I can confirm the above patch () fixes it. (I guess why it doesn't work on my physical laptop is due to some imperative changes in my $HOME.) Would you like to make a PR?\nThis seems to strangely only work in bash, has anyone gotten this to work with ZSH? I suppose I could source the file manually though I'm not sure which path the ${} refers to?\nI have this patch in my tree which works for ZSH: Edit: I see I posted this exact diff above :laughing: Kinda forgot about this, should really subit a PR.\nI've that bit into my config and have done a rebuild, but don't get the desired behavior. Where does the actually output to (which file) - would this not be needed in my .zshrc?\nIt should show up in . The exact patch I posted relies on btw. You could also include it straight in your configuration if you're not using gnome.\nMany thanks - I'm a bit cautious to try that line as I've previously had various issues with fonts / themes / icons. I've found a direct copy of from my old Arch installation and just sourcing that manually for now -works perfectly! In the event it's useful to someone:\nSorry, guess I wasn't clear. I meant including in your , that way it isn't hidden behind the flag.\nI'm struggling to find a way to represent that configuration snippet as a nixos option that can be generalized to just mean or support for vte terminals. For example it could be but vte isn't a program.\nSince the VTE lib comes from , what about something like ? In the same way we have . also have a bunch of shell configuration (, , , ...), so it could make sens.\nYay, great suggestion I can do that :+1:\nMy old solution posted no longer seems to work after upgrading to nixos 20.09. I can't seem to find the actual solution to this, any pointers?\nWhat was your old solution? If you enable gnome-terminal you will have the vte module enabled. If not enable them manually\nI've got: However I still seem to run into the same issue, after running . As for the old solution - that was sourcing the file I had from arch linux. I also tried: And sourcing that file with - that still does not seem to work.\nI managed to fix this by setting to instead of . No idea if this setting perhaps got reset somehow - but it worked previously. :)"}
{"_id":"q-en-nixpkgs-2476f29156f9331e3970536afe515a840aafe8ca1ab6349b2ca2e1755d09ea8d","text":"nixpkgs rev\nI wonder if this is another case of the socket path being too long...\nTests appear to work iff sandboxing is disabled.\nSame issue reproduced locally on\nIf the hydra build failed, how did this get released to the 18.03 stable channel?\nIt is not a release blocking build. Here are the constituents: I'm more curious how it ever worked. I can't really imagine that it hasn't changed since the sandbox was enabled on Hydra.\nI tried some older revisions and it really looks like DBD-Pg tests never worked in a sandbox. The package was last updated 2015. Bumping to current DBD-Pg 3.7.4 doesn't help either - it first results in an error about the socket path being too long, and after patching that the tests freeze at the same point as with 3.5.3. So let's stick to 3.5.3 and just set .\nSounds good to me, will you or shall I?\nI'm working on it."}
{"_id":"q-en-nixpkgs-93b4ebe61026aed47a087b086b806f31e337c1f87dd22898a38adea86369af84","text":"pxelinux (from syslinux) is compiled without support for UEFI PXE boot: I had a quick look at getting it to compile uefi support but I didn't get very far. system: host os: multi-user?: sandbox: version: channels(root): channels(peter): nixpkgs:\nDoes the NixOS netboot image support UEFI boot? I did try netbooting NixOS (18.03) a few weeks ago with a pre-compiled iPXE UEFI loader. It started booting the kernel but failed to mount the initrd (kernel panic no root).\nThe image is using pxelinux.0 instead of lpxelinux.0 it seems and you need lpxelinux for HTTP support so maybe that's what is going on.\nThe http support seemed to worked just fine in my case. I did serve the iPXE UEFI loader via tftp, the kernel image and the initrd via http (initrd is too big to load trouble-free via tftp). The iPXE loader pulled bzImage and initrd via http and began booting the kernel. The kernel then failed to mount the ramdisk for some reason I do not understand. I solved the problem by setting 'Legacy mode' in the firmware (Supermicro) and booting everything in classical BIOS/PXE mode (no UEFI components involved). I can try to reproduce the problem and send you some screenshots if that helps. My vague guess is that the NixOS netboot image needs some modifications to be UEFI capable. I just wanted to raise awareness that there are potential problems with UEFI-PXE down the road.\nIIRC, specifying on the kernel command line is required with EFI booting for some reason (edit: when booting with iPXE, that is).\nwould fix this."}
{"_id":"q-en-nixpkgs-ac75e9ad8a9249427c9d50f6ec3694b636c6c505d833476e5ba7c9d1e0398dc4","text":"When using with and the build fails when trying to write to a in the It works fine if the list of packages is empty or only contains packages that are normally in the environment. Oddly enough, if you run in the it does not produce an error, but it also does not actually include the package, No idea if this is related to system: \"x86_64-darwin\" host os: Darwin Kernel Version 16.7.0, Sierra multi-user?: no version: nix-env (Nix) 1.11.4 nixpkgs: nixpkgs-channels, commit\nI believe the root cause of this problem is that the libraries like , etc are being installed in but the extra packages by are being installed in and the is only looking in one of those locations. I have a work-around hack which I will post when I get a chance.\nCan you confirm that fixes this for you?"}
{"_id":"q-en-nixpkgs-2022f05ea2a5f672cc2a7e0ee96de206001f71f46cfe929a9907a3729f633342","text":"This should be . A similar issue happens for 7.59.\nHa! Yes! Currently this happens: Will submit a PR for this shortly, thanks!"}
{"_id":"q-en-nixpkgs-5e27c3972d659c585e01c92fa4ce6c2e4c3b78e9f52d60689f5195e3c6cb171f","text":"After got merged, and is in the nixos-unstable channel, I noticed after I rebuilt my system that when sddm starts and I type in my password it just goes black and restarts sddm. Meaning that I can never reach the desktop :( I could also not find anything useful in the journal. cc\nIn a VT, try this as root:\nAhh that makes perfect sense and worked :hand: Is there any real way to make sure this doesn't happen for someone else?\nThere was a PR some time ago that was supposed to address this (it was merged) that would simply clear out the directory on start but it looks like that didn't do the trick. I don't remember which. IMHO, the proper way to handle this is something like this:\nI also often have this problem.. could you open a pull request that removes the old solution and that adds your solution?\nThe commit was - I'll PR this."}
{"_id":"q-en-nixpkgs-70b19086e5c4419e9d5e9a75956b5d9bf12c9c0aa22ecbdbbc92390e0c2dce29","text":"Passing to a Haskell package has no effect, the resulting binary is . CC\nI have checked that at the time runs, shows as but shows as .\nThe solution seems to be to pass  in when is given.\nPR with the fix at"}
{"_id":"q-en-nixpkgs-4afd88bcf8df20616ddcebb4b1057e1e163594cf20e08d153d01125912a9b1ce","text":"I expect all of these to not return \"No such file or directory\" and I expect the \"bad interpreter\" to be patched. I think it still \"works\", but compilation output should be clean such that when debugging something in the future there is not first a mess of other warnings to go through. Multiple issues: system: host os: multi-user?: sandbox: version: nixpkgs:\ncc maintainers\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nThe build currently complains that is missing: This program comes from ; I believe needs in .\nI marked this as stale due to inactivity. &rarr;\nAs above, I think this is just a matter of adding the right .\nI marked this as stale due to inactivity. &rarr;"}
{"_id":"q-en-nixpkgs-a5ddcae0bd61ab08184c8e76c2de8bb71d3bef3473761a512a4d98e899f6a475","text":"seem the Keras preprocessing routines were split off into their own package in 2.0. The Nix version was then updated to 2.0 in commit without verifying that it worked. This broke the package as it is now necessary to also package the new package and provide it as a dependency to\nWe do semi-automatic batch upgrades to the Python packages because not all of them are well maintained. This also means we cannot fix all breakage that occurs during these batch upgrades, however we try to fix as much as possible. If you have the time, please package the new dependency and make a pull request.\nThanks for taking care of that."}
{"_id":"q-en-nixpkgs-9a89aad9ddf445b88fc570bea7543b2d6398d1aeb5c41ff02509e947847d5166","text":"Commit introduced type specification for (which was absent before), but this specification is wrong. From within release-18.03, import the tomcat module from the master branch, and create a minimal Tomcat configuration that uses : And see it fails: When using the native module from release-18.03, everything works. system: host os: multi-user?: sandbox: version: channels(paul): channels(root): nixpkgs:\nIt's only semi-related, but why on earth is the type of a list of packages, and not just a list of paths or strings?? It severely limits the configuration options for no apparent gain.\nFor the record, was an issue about overly restrictive type of ."}
{"_id":"q-en-nixpkgs-10c135dd9a3f5fe87794a5656f24cbd677ed770fa281ba52c920bd1bb4efdf2e","text":"Not sure if I'm doing something wrong, but while returns all ghc822 packages, throws an error: Do I need to somehow feed ghc-compact into this? My info:\ngets the listing of packages going again, but the package depending on still don't build, as they have a constrain and ghc 8.0.2 has base 4.9.*"}
{"_id":"q-en-nixpkgs-e563dc5dfb462b5ec08ed7a9c698cf6706184754c31df2ef0639fbfc1edca6c7","text":"Shotcut in theory supports LADSPA plugins. However, LADSPA plugins are not found by Shotcut on NixOS version Install a bunch of LADSPA plugins (by listing them in ) Open Shotcut Add effects to a track Look for audio effects: no LADSPA plugins are shown Please run and paste the results.\ncc maintainers\nI tried to build shotcut and mlt with ladspa-sdk and set LADSPA_PATH, but did not manage to get the ladspa plugins working yet. ,-(\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nfound out why: configure is using , which is not available, and therefore does disable ladspa. Preparing a PR."}
{"_id":"q-en-nixpkgs-60a6ec8c014fe9b34b4316efa9fea5d2ddb30b6ffe6f7b8171fb108ccfc92a15","text":"when trying to build vim with configuration from it fails with: 'gdk/gdkx.h' file not found This issue just started to happen 5 days ago when I updated my nix to 2.1.2, before that it did build this derivation. On Darwin: Try to run in nix repl: system: host os: multi-user?: sandbox: version: channels(oren): nixpkgs:\nSo this only seems to affect the configurable vim, not the regular one? That would suggest that there's a simple compatibility fix that someone made to the main derivation that wasn't applied to configurable. I can't look at it right now but anyone in know what's up with this header?\nDid you override guiSupport or set ?\nmy call to configurablenixpath as gui = \"auto\", if this what you mean.\nYeah that's the problem, gtk is included conditionally depending on what version is requested.\nRunning without it(leaving the default value for the GUI) is still causing the error.\nI'm still having this issue - did you get it working in the end? Would you mind sharing how..?\nI had my vim pkgs set pinned to 1803 and it worked. today I was trying to update the latest nixpkgs to build vim with and I still get this error.\nIn Darwin(MacOS 10.15) I was not able to compile a vim from vim/configurable but for what I was looking for (compile it with python and Lua support) I was able to compile from this is what worked for me:\nWorked around this by disabling the GUI, works well if you don't need it:"}
{"_id":"q-en-nixpkgs-6488de6da0e95559f8e17699723986c81b01685a98075e89bb43f8aa74fdc0f3","text":"If account-daemon restarts in a gnome3 session it totally breaks and becomes unusable. What happens for me if account-daemon restarts is that I see the screen flash like the login animation from gdm. All programs are still up. But the mouse is frozen and I can't even switch to a tty with my keyboard. The only thing that it actually responds to seems to be Ctrl+Alt+Delete which triggers a reboot. And this indicates that I'm actually in a tty but I see an image of my gnome-session. This is a problem when doing nixos-rebuild switch etc. I had a proposed fix in but dropped my branch because I was unsure if that fixed the issue, but I'm having that overriden in my personal config here: And it's much better while doing or or so than it was before. This can be tested by just doing a . system: host os: multi-user?: sandbox: version: channels(etu): channels(root): nixpkgs: cc\nThis is also breaking my ui whenever I try activate a configuration. Stack trace of , with and compiled with . Stack trace of from a coredump, with regular libraries:\nFiled bug upstream:\nFor those looking to work around this, it should be sufficient to remove any config. (Untested).\nThis still an issue\noh, sorry. I've forgotten about this issue and stopped running gnome like 8 months ago. So I have no idea :D\nI'm able to confirm that when using in GDM that restarting the accounts-daemon does crash the session. And isn't a sufficient workaround.\nI'm seeing a very similar stack trace from in my user session after a -triggered . I'm not using GDM auto-login. I don't have debug symbols but here's the stack trace from . Note that the top of the stack is still in in libaccountsservice. When I went looking at the source for these functions, the only thing which made sense was that returned , which led me to the upstream bug and then to here.\nHad the exact same issue today. my backtrace is identical to that of\nThis needs an upstream fix as reported in This is very confirmed.\nI've been testing the workaround on one laptop and not on another. On the laptop without the workaround, upgrading from to caused to crash. On the laptop with the workaround, upgrading to that version from an older version () did not lead to a crash. why did you say this is not a sufficient workaround? It's kind of a pain to test so I can easily believe I'm missing something. Is it important to restart on upgrade? I assume no other distro has reported this bug or there'd hopefully be more activity in the upstream bug report. Are other distros just not restarting it on upgrade for some reason, maybe? Perhaps it's that… because NixOS treats as changed if any of its transitive dependencies change, we upgrade+restart it a lot more frequently than anybody else does? Maybe nobody else is noticing because changes rarely and other distros aren't triggering the crash often enough to annoy people into investigating?\nI'm also seeing this crash and UI hang during . I am not using at all (I'm using ). I'm able to recover the hung session if I can get to a terminal by running . I left a note in with log output.\nStill happens on NixOS 20.03 because the fix was never ported. Here is my backtrace from gdm, which seems to be exactly the same as reported earlier in this thread."}
{"_id":"q-en-nixpkgs-63c5512dcf02b83481258ff1b5c79fb6eac302eedf4948f0808027942144c0f0","text":"Run , observe: Please run and paste the results.\nVery weird... It must have come in from a recent update. You can find libLLDB in the lldb derivation:\n/cc"}
{"_id":"q-en-nixpkgs-40db58e56f6986a2a35929211414ea6ccd1bc05b8da5880428bab813d730f6b6","text":"On running I get: Add to and run. system: host os: multi-user?: sandbox: version: channels(lev): channels(root): nixpkgs:\nWell, the output has at least changed a bit:\nmy pull request fixes the . It broke down because the package had moved to but requires . But according to the browser is discontinued and `firefox-esr-5.2' is marked as unsecure so I would rather advocate for removing this package altogether."}
{"_id":"q-en-nixpkgs-cabb8cac2639eb85be5db5b6b1ee1ae0497beb82f60e66fb3a98f33ab24ea44e","text":"Whenever I try to install both and I get a collision error because both derivations contain : Do files need to be in the packages? The same problem can be reproduced by building a environment where both and are in:\nI have the same problem between and ."}
{"_id":"q-en-nixpkgs-91edeb91b65bad3066995d5346776f004b0a4724f93e127d89ff814d9c3e3bdb","text":"Commit from updated the default version of the GNU Privacy Guard a.k.a. GPG to version 2.1. Unfortunately, this update broke the gpg-related tests of , , apparently because there is some subtle change in the behavior of the GPG 1.x compatibility layer (git-annex relies on rather than using gpg2 directly). To remedy that issue, I fixed to use version 2.0 in . This is no good solution, unfortunately, because now people who rely on `gnupg1' end up having a version of GnuPG 2.0 in their profile that cannot talk to the gpg-agent from version 2.1, which happens to be the one that's installed in the system by default. In it's current state, gpg1 does not work reliably in NIxOS. To work around this issue, I would suggest that we revert to version 2.0 until we've had time to figure out what exactly has changed between versions 2.1 and 2.0 and why this affects .\nFYI: Another thing I came across, due to a broken clawsmail pgp, was that the handling of gpg-agent differs from version 2.0 . GnuPG 2.1 do not use GPGAGENTINFO anymore. Instead it search for the gpg-agent socket in\nFor now at least, setting GPGAGENTINFO to $ seems to make claws-mail work."}
{"_id":"q-en-nixpkgs-d4c86356f8abc140aa2a44302c205ccb3e0bff1cc96d0d42f3a0b95858f49d4c","text":"repstopdf does not seem to be included anywhere in texlive or ghostscript. This means that texlive can't include .eps figures in pdf output.\nThanks!"}
{"_id":"q-en-nixpkgs-99f627f6a5a79c56c7b4e372da226cce32a76e914c95ffad475d02d7d1c13f00","text":"dvc () is version control system for \"biggish\" data and code It integrates with Git but offers a solution for managing and versioning big data sets that come with code commits. Related technology () are like git annex and git lfs. If nobody gets around to this, I may work on this later.\nNeed to first package the Python dependencies nanotime, grandalf and asciimatics dependencies.\nThe extra dependencies: These will need to be packaged as optional dependencies and provided as a separate alias. Something like and or something like that. What are some good suggestions for names? The and don't appear to be packaged.\nGoogle cloud storage is packaged already here: However is still not packaged as is though. The pypi is:\nInitial PR for azure-storage:\nInital PR for nanotime:\nInitial PR for grandalf:\nInitial PR for asciimatics:\nFor the smaller dependent packages that are unrelated it would have been preferred to group them in one pr.\nHowever too late now :smile: so just to know in the future.\nYea that would have been better as I could have then worked on dvc commit as well, but now that all dependent packages are separated between different branches, that's not possible. So I'm just waiting for those to be merged to master.\nInitial PR for dvc:"}
{"_id":"q-en-nixpkgs-0e13d9ded31313ce69f069abeca7dda61bc291231da513ca6bb48c0201e76015","text":"Anyone want to collaborate? I will try my hand at packaging it in a bit, and will probably add it to the wayland/windowmanager overlay I'm maintaining along with a nixpkgs PR if I get it working. Arcan \"Arcan - [Display Server, Multimedia Framework, Game Engine] -\"Desktop Engine\" http://arcan-\" Durden \"Desktop Environment for Arcan \"\nI've succeeded in building and running Arcan and Durden. works but I'm sure there are things in the closure that don't belong (X11?). I started with Void Linux's dependency list and kept adding things cmake complained about until it built. Getting it running was challenging because it depends on a bunch of environment variables being set correctly and neither the error messages or the documentation are all that clear. It also wants its binary to be SUID. Here's what I've been doing while working on the package and figuring out how to run the binary: From a virtual terminal: Here are my settings for the environment variables: I created all the directories in $HOME which are referenced in the variables. I don't have a package for Durden. I installed it by cloning it off github and then making a symbolic link to its subdirectory from . I'm running this on a laptop with Intel graphics, and have been working on it in a virtual terminal with my x session left running in another. The one time I switched back to xserver without killing arcan first, it hung my laptop.\nThat is great to hear thank you for sharing. I'm pretty occupied right now with some traveling, but I'll try to take a look at this soon. Thank you for writing it up though, I'm sure someone will stumble on that and be thankful too! EDIT: oops, I thought this was on a different repo, I probably would've worded this differently otherwise. Either way, still excited to see this. :)\nI've and am typing this in Firefox running in an X server in a window on Arcan. The package definition builds an executable called and you can invoke it from a shell prompt in a Durden window to run it. To run a second instance, use . To start an application or window manager in the X server from a shell prompt, set first as in . It doesn't automatically detect when its window size changes, but works. The font path part of the derivation isn't right. I made two attempts at making it work and left them both in the derivation. The code in puts a font path into the configure flag provided for that purpose, but the resulting executable can't find the fonts. The other chunk of code in links fonts into the derivation, and that part gets the truetype fonts to work but not the bitmap fonts. I'm happy to run so I stopped there, but someone who knows more about X fonts than me should take a look.\nI was curious about the various Arcan tools (clipboard, image viewer, network proxy, wayland bridge) so I packaged them. There had been a lot of change in the tools directory between the last release and master, so I packaged the tools from master. So the latest commit in adds , and several child packages under . I haven't figured out how to get the Wayland bridge to work, but I wasn't using Wayland before this so may be missing something obvious. I have been running and with Durden on my laptop for about a week now and am finding it very usable and stable.\nHey there and Wish I had found this thread earlier. I had a go at packaging Arcan at . It builds, however it still has some rough edges at runtime (most glaringly, mouse and keyboard don't work if not running as root - addressing that now). I'm going to study what the two of you have discovered so far, and maybe we can join forces in bringing this very interesting piece of software to the Nix ecosystem.\nHey I've just updated with a commit to make it build more recent versions of and . I've had no problems with but the recent litters the screen with mouse cursor trails. It would be great to get it running not as root so keep us posted on your progress.\nTo get the keyboard and mouse to work when not running as root, just add the user running to the group.\nthanks, that worked! I brushed the dust off my Arcan/Nix code just to try out your suggestion. Now I have Arcan, its core utilities and the example compositors (except which fails to find something or other) running on my system, including a terminal - one more thing to run besides , , or in a vanilla console! Neat! I skipped the whole overlays thing and ended up with a file that can be submitted to Nixpkgs (or imported into ). Guess I should share that part of my config as a gist or something before submitting a PR? Or is there something missing, as might be suggested by Food for thought a.k.a. \"and then what?\": Port my i3 keys to a fork of Durden, ripping some features from Prio to finally organize my daily terminal-browser-devtools grind in a sane way :grin: We have a powerful engine on our hands and a whole lot of Lua ricing to do. Would be cool if some logic from can be ported over to an environment outside X.\nGist: (Sure is cool how Durden and Prio have different mouse capture behavior btw)\nI marked this as stale due to inactivity. &rarr;\nThere is still some good progress over at the gist.\nI found this while looking into a new project (, ) posted to today. (By same developers as arcan, and uses arcan, so I was curious if we had it packaged.) Just cross-referencing on the off-chance it helps anyone here. I'll also add this to the gist."}
{"_id":"q-en-nixpkgs-1395b67f9c4ccc3dcce523082d466dc4f4cdb3fb378dcb5b6cfdad9288789066","text":"Data that was once config and recorded on doesn't get updated when the configuration changes. This is a violation of purity. And more than that, it's very surprising when trying to install, hesitating between pgsql and sqlite, and struggling before understanding that the configuration doesn't get used anyway :p cc\nthat's a bit tough. We have a unit, that's supposed to do the installation. We use the command to install nextcloud (passing the configured database and adminpass settings as command parameters). This CLI generates , and installs nextcloud. The script checks the presence of , and skips installation on following starts. If you now change configuration from the module, we won't run the install command again, as the config file already exists, and assumes nextcloud is already installed.\nHmm… Maybe it would be possible to generate the somewhere in the store, and symlink to it? This way the next time runs it can in addition check whether points to the correct store path. Now, I don't really know what exactly does, so maybe it's not actually possible to do this because eg. one would need to re-run on certain configuration changes? In which case it'd be hard :/ but at least we can maybe make fail or something like that, so that the issue is at least noisy?\nI don't recommend to do this, I just tried that (after upgrading one of my personal systems) and this tries to recreate a database schema which can end in very bad results. Unfortunately I doubt that this will work properly, nextcloud wants to merge this during the runtime. E.g. it adds a after a successful or adds trusted domains when running the corresponding command. I'm using this as well now and hitting these issues, unfortunately I don't have useful ideas either ATM, but I'm happy to discuss better solutions :)\nNextcloud edits it's in a lot of different places, so having it generated through Nix currently isn't possible. digs into that already, I a comments about our use case here too.\nto avoid other people running into the reconfiguration issue for now, what about documenting that, maybe in ?\nThat's certainly a good thing to do. If nobody else wants to take over, I'd volunteer to do this during the next week as investing time on improving the nextcloud module is one item on my todo list :) I was aware of being stateful, but thanks a lot for linking the corresponding ticket! I'll have a look at this as soon as I have sufficient time to :)\nIn the light of this issue, what's the correct way to upgrade nextcloud to a new major version? Do I just pull a newer , upgrade my system and hope that the systemd job will do the right thing?\nThe service runs an that should apply all changes like database migrations needed for the new Nextcloud version (beware, downgrades aren't possible). As long as you let Nextcloud do its job, you shouldn't get that much issues. IMHO the actual problem here is the configuration of the software within Nix(OS).\njust a simple question (could not come to a conclusion after the lengthy thread) but I recently updated my nextcloud via nixops deploy and the update disabled the calendar module. Did I mess up or is that expected ? any way to configure this from nix ?\nAssuming that you installed the calendar module via the UI, it is installed in . The beginning of my config reads: So in principle nextcloud can edit the apps installed via the web interface (otherwise you couldn't install them via the interface in the first place).\nI don't think there's anything actionable here - the nextcloud nixos module is limited by how Nextcloud (imperatively) manages its configuration. The current state is properly documented in , and we raised our concerns to upstream in - closing this issue, feel free to reopen if there's more to be done.\nI think one actionable item would be to have the nextcloud service fail with an explicit error if the configuration being used is not compatible with the configuration defined in the .nix. At least the user would be warned in practice that configuration is not really applied, and not only by the manual that they are relatively unlikely to actually read. Reopening for now, feel free to reclose if you think that's not a reasonable “solution” :)\nThe problem is we don't know whether the configuration on disk containing database connection information () differs from the one that would be generated through an installation via - we don't render the configuration on our own, but the installer writes it out as a side-effect. If I'm not missing something obvious, checking configuration changes would translate to parsing/evaluating on every nextcloud startup as an , comparing every \"interesting\" key with the one provided through nixos configuration, and eventually then bailing out if they differ. I'm not sure if that's a road we'd like to go and keep up to date, while the underlying bug is a missing distinction between configuration and state upstream…\nSince upstream mixes state and config it's questionable whether their configuration format is somehow supposed to be backwards-compatible (as it's only supposed to be used internally by Nextcloud). Hence I'm afraid that this will make i.e. upgrades even harder in case of configuration changes, so :-1: from me. As the issues are documented in the manual now, folks are warned and I think that this is a better option in contrast to adding more complexity to . I suggest we either close this for the reasons already stated or wait for an appropriate upstream fix.\nI really don't feel good having stuff in that's actually a disguised command line… :/ Something that would sound nicer to me would be an executable that the user would have to manually run, and a bare-bones module that does not handle setup or configuration and just makes sure that the systemd unit is setup. At least the error is obvious (enabling the module without having run the executable would fail), all module options actually mean something, and the impure operation is made explicit. Either that or, similar to , rename all the options so it's explicit that they're used only for initial configuration. Would one of these solutions make sense to you?\nMe neither, but currently we don't have much choice… I really like the fact that we can set up the Nextcloud instance in an unattended fashion, without having to worry about manual setting things up. I really don't want to end with more imperative setup steps than necessary. In theory, we could move some of the options in to to make it more obvious what isn't applied continuously. But TBH, I'd rather like to see the state/config mixing fixed upstream than having to introduce more mental complexity here, especially if this gets fixed and we end up with both incarnations. Really, this issue only occurs while a user changes nextcloud database parameters (mostly while tweaking his initial setup), and not having read the module documentation, where the behaviour is documented. What about adding the following (subtile) log statement, to point users to the docs when they peek into the logs while things are not working as expected?\nI agree that the log statement is the minimum we can do, and would at least give a hint about the way in which things are broken. Now, I guess the reason we're kind of disagreeing on the path forward is that you're much more hopeful than me that nextcloud will ever fix its stuff -- I'm assuming it'll get fixed in years at best (and even then…) -- and if it's in years then I'd much rather live with for years then spend a year deprecating it by making it conflict with new options than try to live with from now on and have users fail to understand what is going on. Now urhm, I'm investigating the module a bit deeper, and… is there any reason we can't put the database configuration etc. in ? We could then refresh on each activation, and it should be enough? If we have a way to couple it with something that automatically runs iff the database is not ready yet then the setup would actually be… almost perfect?\nAfter having a deeper look at the sources ( at github, (6 commits ahead of v16.0.0RC1)) I guess that this should be possible (I didn't test it yet though). Configs will be loaded from using PHP's function[1] while all files ending with will be taken into account. Hence works as all config arrays will be merged together. This config will be loaded at [2] and is available as which is then used in the DI container[3]. If I understand the code correctly, I assume that it should be possible to render all config options in as config files found by glob will override values from [4]. [4]\nNice catch! This might indeed be possible - we should make very sure that config overrides work the way we think they will - seems to write the merged configuration, and not just the overrides (and merging this with other config in the wrong order can lead to wonderfully weird to debug issues) ;-) While I'd still prefer nextcloud splitting config and state properly, and being able to cope with read-only (base) config - it might be a direction worth exploring. What do you think?\nWell, for the record it was actually idea :sweat_smile: I guess so, all the stuff from is also in my at the moment. I'll probably have sufficient time next weekend to check if I can trigger such a case. I can't promise that, but unless anybody beats me, I can hopefully prepare a fix for this :)\nDid you gain some new insights on this?\nUnfortunately not... I was pretty busy back then. But as this only requires some fiddling in the module and no patching of the actual package, the change shouldn't be too painful, so if you'd like to take over... :sweat_smile:\nthere's with a fix now :)"}
{"_id":"q-en-nixpkgs-0086b89fbe9f0e12e14b4dfb64dac30b9e01275681f7507b7cce73faf898fb6e","text":"Our current legacy nvidia driver of the 340 branch does not work with kernel version 4.19.2: NixOS is currently on the 340.104 version, but upstream is on 340.107: I hope that updating to 340.107 resolves the issue, but I don't understand how to update the different checksums for the nvidia driver. /nix/store/kic17fw8wil74k04kzh3dha43izrr9j3-bash-interactive-4.4-p23-dev copying path '/nix/store/kic17fw8wil74k04kzh3dha43izrr9j3-bash-interactive-4.4-p23-dev' from ''... system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nThe hashes are located at and can be generated like so: Please try:\nThanks alot for giving me a hand and teaching me how to get those four hashes! :+1: Updating to 340.107 indeed resolves this issue. PR:"}
{"_id":"q-en-nixpkgs-23ed33c89816a88ac8268cf1f1a4974bf3f75fa0932454b61f0fba4cce28bb46","text":"During configure, ’s configuration script checks for certain features of the command, to quote the source: The latter two are not available on macOS, but nevertheless get enabled (I suppose is in when configuring?). Things we can do: with on darwin. haven’t really looked into how this build system works, but maybe we can pass configure flags to explicitly disable these features no matter what on darwin? it away.\nFor searchability, a typical error would be:\nFTR I’m not quire sure where to apply any of the fixes, probably in . (/cc"}
{"_id":"q-en-nixpkgs-935f8eb36a9b85ac42e9d47f23589af752ca7edba7b3a7cdd62a5a359fb7ec33","text":"(this is a qt specific test): Loc: [t-(458)] PASS : TofuInfoTest::cleanupTestCase() Totals: 7 passed, 1 failed, 0 skipped, 0 blacklisted, 1047ms This seems to affect master too, but for some reason it's been built successfully once. The aarch64 build started failing when updating qt5 from 5.11 to 5.12 making it the likely root cause: Edit: the cause was expired testing keys: cc maintainer cc as you've been active in keeping plasma/kde up to date lately. ping staging next:\nDid you found ? Sounds very similar and should be fixed...\nNo didn't find that, but yeah, that sure looks exactly like the same issue. I'm guessing the qt issue were never actually fixed: Blame on the test gives no change at least:\nAnd applying the suggested patch does fix the test: diff +++ gpgme-1.10.0/lang/qt/tests/t- -455,7 +455,7 private Q_SLOTS: QVERIFY(result.numSignatures() == 1); auto sig = result.signatures()[0]; QVERIFY(sig.validity() == Signature::Marginal); QVERIFY(sig.validity() == Signature::Marginal sig.validity() == Signature::Unknown); auto stats = ().userID(0).tofuInfo(); QVERIFY(!stats.isNull());\nSo, what are we want to do? Do you maybe want to comment on the upstream issue?\nHmm, strange, I'm getting the same failure on master (using ), forcing single core does not fix the problem though, so probably not a race condition. I've reported the failure as you suggested:\nThe cause is most likely qt 5.11 to 5.12 upgrade, as the aarch64 build started failing the test then:\nSo the keys used in this test expired on 2019-01-06, which explains why we suddenly saw this without any apparent reason. :\nThanks for fixing this and sorry for my late reply... (Edit: And of course also thanks"}
{"_id":"q-en-nixpkgs-bb516e53b080a4db41f64e5eaf8093870659249789151e0937a681f573d4ff23","text":"An update of zip-archive package from 0.3.3 to 0.4 in is causing a runtime error in elm compiler (). It would be good to set the elm's dependency on zip-archive to version < 0.4 until this gets fixed.\nprobably best to patch it\nI think fixes it, will try to find more time to test tomorrow morning.\nShoot, I forgot all about this -- I did confirm that commit fixes it, I will clean up and open a PR soon."}
{"_id":"q-en-nixpkgs-a1712b790e9ae1f174d7743c328765096203ba95a632d4d644070f31aaef919f","text":"should be enables horizontal scrolling\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on ."}
{"_id":"q-en-nixpkgs-4d2f728f6cb5ebf00e05dacc5dc43308462d064fdf8889d7e7a4e56906e78ddc","text":"I tried to enable syncing Unix passwords to Samba via PAM using the option. However, I see a bunch of errors in the logs because the library PAM is looking for, , doesn't exist. Use the following config: Please run and paste the results. system: host os: multi-user?: sandbox: version: channels(USER): channels(root): nixpkgs:\nSame issue here. The folder it was searching for () only contains on my system. Additionally, querying for presents no results.\nJust experienced exactly same issue on 19.03. Is samba broken in this release?\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nIt's still relevant now. I noticed this issue in my system logs as well. Samba is not broken, they've just deprecated this feature. See and . Here's a PR for us:"}
{"_id":"q-en-nixpkgs-5661c8290e95dc14a3c79a8f0e2cf9ede6c97db9d2179548daf38c91205c4211","text":"always has the same , and the same filename inside. Build NixOS with , look into I think it would be sometimes more convenient if the kernel version would be clear for each (which is actually initramfs). I think it would be better to literally include the as a suffix to .\nThis sounds like a sane idea to me."}
{"_id":"q-en-nixpkgs-9d4e75c00aaf57b6624d5f35dd439fb6654cd99c7f69bac0f98c4b321c44ab24","text":"I have been burned by this several times; and I am not alone: When calling with multiple arguments, we should either patch all of them or print an error. Ignoring arguments silently should be a crime. cc"}
{"_id":"q-en-nixpkgs-88dd3d078278ef086186d885142a0f3b67bfbc9201d49256ef50833f60798ae8","text":"building some cross-compiled packages and inspecting their run-time closures: will show that depends on (on master ). this is because the package isn't spliced correctly: The cross compiled packages depend on the folder, but this path also contains x86 binaries in it! (like ). This causes a bunch of other run-time x86 dependencies to be packaged with the cross-compiled binaries. pointed me here: , which is possibly broken. Trying to move the ARM libs into the output caused a cyclic dependency between the lib output and the normal output :( so I'm stuck & posting this issue. build a cross-compiled binary with a cross , notice that it has x86 binaries in it's run-time closure (assuming the build system is x86): system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:"}
{"_id":"q-en-nixpkgs-b48f175e0b2d737ebe3d2d904c38db99d924d0393e7582f4b88e90fcea6e0989","text":"fails with the vdirsyncer version from nixos-unstable. it works if I install vdirsyncer with pip. Output of :\nWhich do you get via ? Is it the old 0.16 branch or the new 0.17 one (which was mostly rewritten in rust)? We have the latest git version of 0.17 in nixpkgs. Unfortunately, so it might be some time before this bug is fixed.\noutputs this:\nSame problem here. Was there any reason to go for an alpha release except for the issue referenced in ? If not, a downgrade to 0.16.7 should be fine, since the original bug is still open and probably won't be fixed for a while without anyone maintaining vdirsyncer.\nThere are , so I'd rather not downgrade. I'd be happy with a package for the 0.16 branch though.\nThat'd be great! I haven't looked into packaging yet, so it would take some time if I tried. Would you mind helping out?\nI just opened a PR. It’s my first time contributing to , so I hope it’s up to snuff."}
{"_id":"q-en-nixpkgs-accdbdce0560d041d95bad603092867c4810d2bc49437b18f14c393c69140a4e","text":"After in nixops, we can remove the locally pinned version in nixpkgs.\nNot yet, I would like to keep those for building nixops v1.6.1 for any users of the Azure backend.\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on ."}
{"_id":"q-en-nixpkgs-7fe6a8e6f47334f8a12f447f6410809b0efcaf526d608e4f89c91539ca5e7e6b","text":"It should always show the NixOS version. Reproducible on: Lenovo Thinkpad X230 with grub Not reproducible on: Apple iMac 11,3 with systemd-boot (EFI) Thinkpad L470 with systemd-boot (EFI) system down system system: host os: multi-user?: sandbox: version: channels(davidak): channels(root): nixpkgs:\nI can second this. It's been the case for ages.\nYes, I'm not sure why this is. It used to show it at startup, and I don't think anything much has changed there.\nI did an update yesterday and bootet the system today and the version was shown. It's a Thinkpad L470. I will see if it fixes it also on the system i used when reporting this issue. system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nDoesn't work here on latest unstable.\nSeems to be a problem specifically with GRUB.\nI'm using systemd-boot, not GRUB.\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nProblem persists.\nThis is because the Plymouth plugin is missing from the initrd, along with a font with which to render the text. See ."}
{"_id":"q-en-nixpkgs-cc89dd77666c5347ac72eaa42f07c213cd4aabf2560891256a2ae6837727730a","text":"The rust crate \"built\" assumes that certain environment variables are always set -- to an empty string if necessary. This is what cargo does, so let's emulate that. See for more details. See I am preparing a patch to fix this, so I skip this."}
{"_id":"q-en-nixpkgs-db1f7ca45c6d1fb1953e6b889d94bb1f3d3ce4dbc8b1c7297a07429d0a198ceb","text":"This expression fails to build on 19.03: The problem is that its checkPhase is not correct as the command cannot be found. I believe that should be in the for it to succeed. Run the below inside The error is: Please run and paste the results.\nIt can be put in as is usual for pytest because we have multiple versions of it within . However, the tests fail:\nNot sure what is causing that error, but I use it fine for my purposes.\nHow do you use it if it doesn't build? Would you mind making a PR with the fix?\nIt does build, I use a custom pytest- because I was using it before asking it for inclusion in nixpkgs. Anyway it's now being addressed in:"}
{"_id":"q-en-nixpkgs-b94ef09a0193cebe48dc846344cad8627f1f1253479670bfbe50968d5076df96","text":"build of gpgme-1.13.0 fails on macosx . . . nix-env -i gpgme Please run and paste the results.\nI can reproduce this on , but it builds fine on current master\nClosing since seems to be fixed after , but feel free to reopen if you continue having issues\ni am seeing this again on\nbisecting suggests it started with or (in the latter commit, indent fails to build, rather than getting the File name too long error, though)\nSame here (seeing this again)\nIt seems like this is caused by a of unix socket path length limit (which is just 104 on darwin). A workaround that works for me is to build it with a different TMPDIR:\nRelated to\nThanks for the workaround using it, gpgme install gets stuck at\nI had the same problem. I think this might be an unrelated issue. It did build successfully on the second attempt for me.\nIt would be nice if somebody could investigate how to disable just the problematic tests instead of completely turning them off with .\nTurning off the tests didn’t work for me, it seemed to fail on the build step\nwe could add a patch to disable these tests on darwin:"}
{"_id":"q-en-nixpkgs-73fc1a1898a192f097af5a68c6cdbb43449b126f46f4114841b87685c1234ad3","text":"application can't find help files Install from nixos release channel, go to main application screen, click on Help menu and select either \"tutorial and concepts guide\" or \"contents\". Application complains with a modal reporting \"GnuCash could not find the files for the help documentation. This is likely because 'gnucash-docs' package is not installed.\" system: host os: multi-user?: sandbox: version: channels(root): channels(zarel): nixpkgs:\nAs a workaround: the very same documentation is available online at this address:\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nstill happening\nPinging maintainers\nThere is more to it. Gnucash installs its document files into plain without subdirectory. Because of that common files like do have collisions with analogous files in other packages. Edit: Ah, that's another issue. Which was resolved in\nDocumentation for Gnucash is a separate project: It should be separately build and installed in order for help files to be available.\nI marked this as stale due to inactivity. &rarr;\nI've built the documentation and put the documentation in the XDGDATADIRS. I can browse the manual and the tutorial via but running gnucash in debug mode says \"Unsupported action\" and according to strace the application doesn't try to open the documentation directories... and the same happens for evince (can see the documentation in yelp but the application won't display it), maybe it's the same issue\nNot stale, I'm still having this issue\nI marked this as stale due to inactivity. &rarr;\nissue remains.\nThank you for the fix! Do you think is there a way to warn the user about not having yelp in a profile?\nI don't think there are any convenient/good ways to tell the user about it. At best, we could add it to the description?\nIf gnucash documentation depends on yelp, would it make sense to express this dependency in the gnucash derivation?\nApparently, it must also be in MIME. This sort of stupid behaviour looks like an upstream bug to me.\nIf an application shipped its help files as HTML and required you to have a web browser installed to view them, I think it would be silly to add a dependency on any specific web browser. Likewise with help files - if (at least theoretically) Yelp is just one of several possible help file viewers, I don't think it makes sense to add a dependency on it. A better indication of what's wrong (\"you need to install a help viewer\" rather than \"couldn't find the docs\") would be nice, though, which I presume is only really doable upstream."}
{"_id":"q-en-nixpkgs-fb489cab2ecb549d280a05ca3463755f79c27dfa30e3a88a7ea0ffb106f272b4","text":"On today's master () - and probably others - building the attribute of many package produces an output that differs from the output on another architecture. Nix will display a hash mismatch. I tried with x8664-linux and aarch64-linux. Execute this on aarch64 and x8664. The aarch64 build produces the below output while x8664 is happy with the hash in nixpkgs. Please run and paste the results. system: host os: multi-user?: sandbox: version: nixpkgs: and system: host os: multi-user?: sandbox: version: channels(root): channels(andi): nixpkgs:\nx8664 build produces same output for me: system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nI noticed something similar on macOS in Could some recent change have caused this? Perhaps ?\nI reverted in a branch on my fork. My x86_64-linux laptop doesn't complain about the hash sum - as before. The aarch64-linux community box also doesn't complain anymore. Can anyone else confirm this?\nI just made a diff of : The left side is aarch64, the right side is x86_64. Both are done on the same git revision ().\nWhat a weird behavior that is missing.\nProbably related:\nI think this may be related: I also see different hashes on NixOS 19.03 vs. stable. I use the branch, but when CI builds the derivation on 19.03, the hashes do not match. If I build on NixOS 19.03 with from master, the hashes do match. This is quite annoying, because it makes it impossible to have package sets that work on NixOS stable and unstable without using rustPlatform from a nixpkgs newer than 19.03. I'll try to revert and see if I get the same hashes on 19.03 and master.\nThat's indeed it. When I revert , I get the same as 19.03. Maybe this change should be reverted? Since NixOS is not a rolling distribution and people run both stable and unstable, it would really be nice if package sets and overlays work on both versions. This also makes backporting changes easier.\nDiff between the vendored dependencies: Here the differences also seem to be missing dotfiles (e.g. and ). I think this is the upstream PR that changed this behavior: (Edit: oops, I just saw that already mentioned the associated commit.)\nis a critical update and reverting it will break multiple Rust packages. Rather I think we should get that change pushed into the channel, and then update any Rust packages whose changes as a result.\nThe only bummer is that it also breaks people's own packages on the channel as well.\nYeah, it would be bad to break people's derivations as an update to a stable release. At any rate, the 19.09 release notes should probably have a blurb that points out that hashes will change due to changes, so that people who update from 19.03 to 19.09 are prepared. There are probably also derivations in that have incorrect hashes now, since the file that is generated by is now excluded vendored dependencies.\nI’ve updated the title of the issue to better reflect what I think we now know. Additionally, I’ve opened to update the hashes for all packages that haven’t been updated in master yet. I would appreciate a review on it. Once that’s merged, I can do the same thing for the affected release branch.\nLooks like I misunderstood and this hasn’t made its was into a release branch yet, so we’re almost done here. Somebody just needs to add the release note for 19.09.\nwith the changelog PR () being merged, is there anything else missing?\nDon't think so. :) Somebody can shout if I'm wrong."}
{"_id":"q-en-nixpkgs-3497c699f5a5537f9a984ba8289123ed2f7471d14d1ac7d8ab5866c3597822a7","text":". The short of it is that doesn't have a flag on OS X.\nThe in use is from .\nIt seems that removing from solves this issue. I'll drop that in a PR to continue the discussion."}
{"_id":"q-en-nixpkgs-7a497fd8de9d8925321c3ebfbb73a1a63d55d6f7b56dad6b5bb4b7e10d9b762a","text":"Right now if you update the unstable tree and install ddclient the shebang patch is not applied properly and the first line of the script starts with: This reference doesn't exist so the actual binary cannot be started. ) Install ddclient ) Attempt to start ddclient"}
{"_id":"q-en-nixpkgs-4e4360d7d771c75a5d156e2e46606b0fa80e1a351de6b4079ae81f13f3d9607f","text":"In there was an option introduced that indefinitely tries to restart the wireguard service until all peer hostnames are resolved. Instead it should use as introduced by instead. This feature will be available in the next snapshot release of wireguard. cc\nAlright 0.0. has been released for the purpose of this disaster being fixed. care to cleanup the mess?\nI fail to see what mess/disaster you're refering to."}
{"_id":"q-en-nixpkgs-29b367b6930ac438f385974e019914924836038c6782d4b997e87ef2505106ef","text":"The is from over 3 months ago. Some background info: Previously it was updated automatically by bot that runs . But the last builds have failed, e.g.:"}
{"_id":"q-en-nixpkgs-d92827373bfe387ef994e93df2dcdf5cf5aeef07540b1dac1fc8b4969909dc51","text":"aborts and produces the following error: I suspect it's due to , but when I started building a NixOS vm with the merge commit checked out, Nix started building ; I just don't have the computing resources to build from such a starting point. What I do know is that the problem exists in but not in\nIt's indeed likely due to that PR. As mentioned in the message, its up to the maintainers/community to fix the usage of the wrapper in the various packages.\nYep. I have a fix for the \"Qt platform plugin\" error, but the missing Qtml module has stumped me. I've -ed my way toward the culprit, but I'm not there yet; Compiling LLVM, Qt, PyQt... takes a LONG time on my computer ;)\nAccording to git bisect, the first bad commit is This implies that somewhere within that commit the PyQt build changed such that the Qtml module is no longer being built.\nCan confirm the issue.\nI dug into this further yesterday, and though I don't have a solution, the problem points to three things: package needs to be modified to use . That part is easy, and I already have a commit for it. is building in such a way that not all of its modules are included, namely the QtQml module. This results in another runtime error; visible after has been applied. reason is building with missing modules is because during the build process it uses tooling to determine whether said modules should be built. In the past, the tooling would return in the affirmative, causing the module to be build. However, as of late the tooling returns in the negative, thus the modules don't get built. This can be seen in the trunk/master python37Packages.pyqt5 builds on Hydra; just search for QtQml. It is number 3 above where I'm stuck. It seems like it's due to a change in , but I don't know what that change could be, nor how to test for it. Note: I ruled out the mismatch between PyQt 5.11 and Qt 5.22 as a potential cause by forcing PyQt 5.11 to build against Qt 5.11.\nYeah, I've been at least aware that we needed to update PyQt I think is working on this.\nFix in\nUnfortunately, did not fix it. I: out a vm with qutebrowser within the vm The error is the same as I originally reported. I noticed that adds as a dependency to qutebrowser, but doesn't utilize nor . In contrast, my workaround is to --after applying a hack to PyQt 5.11 -- change qutebrowser as follows:\nI think the cases were working for me because my environment has the right plugins. But indeed, wrapping should fix this."}
{"_id":"q-en-nixpkgs-3df482cf471117acd4f5a55e9c8d0e359fa21f311be93f65cf2b902b0a820ac9","text":"Currently python packages in propagatedBuildInputs are included into PYTHONPATH and are referenced by a file in the output (propagated-build-native-inputs), which makes them run-time dependencies: For ELF programs nix can figure out the run-time dependencies by scanning the executables, this fails for python as the dependencies are not referenced in any file. DISCUSSION: pth files could include such references and would handle PYTHONPATH manipulation a nix-support/python-build-inputs could list python buildInputs, turning them into run-time dependencies and further they would be used in the above _addToPythonPath\npth files it is"}
{"_id":"q-en-nixpkgs-c3b29f84eecd6e4476fa212fb223c98c1b48755eccdc987aac494b0467b497f4","text":"Describe the bug qutebrowser doesn't start To Reproduce Steps to reproduce the behavior: the latest nixos-unstable qutebrowser Expected behavior qutebrowser starts Screenshots ! Metadata system: host os: multi-user?: sandbox: version: channels(philipp): channels(root): nixpkgs: Maintainer information:\nThe issue is present since I will figure out what commit from staging it was, but for now my computer is busy compiling :)\nmaintainers:\nRelated to\nAnd\nThanks to for this valuable piece of information: Adding wrapQtAppsHook only address plugins, such as xcb, not loading. It does not address the missing PyQt5.QtQml module. According to a working build of qutebrowser, that module is actually missing from PyQt5. This can be seen by comparing the PyQt5 referenced by working and non-working builds of qutebrowser. In the working build you'll see the files and in `/nix/store/PyQt/lib/python*/site-packages/PyQt5/. In the non-working build those two files will not be present. In fact, I recall seeing other missing files, such as QtWebEngine. I tried some stuff and it seems like since the propagatedBuildInputs are somehow ignored and can not find webengine or qml, even when adding qtdeclarative to propagatedBuildInputs. I have no idea why, but only when I put them in nativeBuildInputs, can pick up those dependency and will produce a pyqt version with webengine and qml modules and qutebrowser will start successfully. I know this is wrong, so maybe someone can find the cause of this and a proper solution.\nfixes this, but it is not a proper fix.\nFix in\nDoes anyone know a way to work around this bug, and get a working Qutebrowser, without having to compile it?\nYou can install a qutebrowser version from an old nixpkgs revision to your user profile. If you have cloned nixpkgs somewhere:\nCool, thanks. But doesn't that build the whole thing? Building a web browser like this takes several hours, no?\nThis package should have a binary cache now.\nNot really: the engine (QtWebEngine) does, but the browser itself is just a relatively small collections of python scripts, so it's pretty fast. QtWebEngine, being part of the base Qt package set should always be cached, even if you modify qutebrowser.\nWhen I installed the version mentioned above, it was still cached. Usually old packages are removed from the cache quite a while after they were replaced. But as worldofpeace mentioned in current unstable qutebrowser should work just fine again.\nCool, I didn't know about that. Thanks!"}
{"_id":"q-en-nixpkgs-c53210a942c675c3cb05e355041290abdee65246610ada59fe0e233647ef6b50","text":"Describe the bug grpc doesn't build on macOS. To Reproduce Steps to reproduce the behavior: -shell -p grpc Expected behavior Should have built without errors. Additional context Build log here: Metadata system: host os: multi-user?: sandbox: version: channels(kaali): nixpkgs: Maintainer information:\ngrpc appears to be pretty out of date , so I'm suspecting that updating all of the pieces would resolve this.\nSeems to fail in CI even with an updated version:\nThat particular package requires manual intervention, as that PR was automatically generated by a bot."}
{"_id":"q-en-nixpkgs-57c47e19974617a9cca4c65aa01ef61fb5b982a24aa832a7424faef73d7c72cf","text":"write python expressions (5.2 in nixpkgs manual) use virtualenv create nix python environments\nDoes this documentation exist? Should it be completed before the release?"}
{"_id":"q-en-nixpkgs-1c8eca7c3b41bb621b5a65b7bcdd4d96173d8d4a531a70e598f2c673c72be104","text":", (nixos-unstable) Scanned versions: nixos-unstable: . May contain false positives.\n` will take some effort."}
{"_id":"q-en-nixpkgs-80cf0e8e33572cd09c7196f20d8dca5546e1ac345e0b0f03c4e587a2552a7364","text":"Project description Imageboard/booru downloader which can download thousands of images from multiple boorus very easily. Metadata homepage URL: source URL: license: Apache-2.0 platforms: Windows, Mac, and Linux\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . on the on .\nStill not packaged for NixOS. So still important (at least to me). I'm trying to package it myself, but it doesn't work due to dependencies on packages from NPM.\nGoing to take a look at this after I open a PR for ."}
{"_id":"q-en-nixpkgs-a149e95ec671c436aee43b025e494b4a29635e145501065056983528e3cb5b73","text":"Describe the bug In mpv, CUDA-accelerated playback worked a few months ago after landed, but now no longer does. To Reproduce Steps to reproduce the behavior: NixOS master with an NVIDIA card and and observe: Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(at): nixpkgs: Maintainer information:\nI will see if doing helps"}
{"_id":"q-en-nixpkgs-f8a265b3b6022fb666ecb533d34f334f3f8137e8bbf24116d59a211d695dc6b9","text":"Describe the bug A clear and concise description of what the bug is. When running dolphin on plasma5, getting this error: /cc To Reproduce Steps to reproduce the behavior: dolphin \"Search for -Document\" error Metadata Please run and paste the result. system: host os: multi-user?: sandbox: version: channels(root): channels(mbauer): nixpkgs: Maintainer information:\nIt looks like adding to systemPackage fixes it.\ncc\nThere is also a similar bug with Kmail if you don't include Akonadi in system path. Something must have changed in the way binaries are found.\nWe should make sure that puts in ; actually, anything that provides a should."}
{"_id":"q-en-nixpkgs-fbd6298c62ff3257c83c9acb25ba6f414e1a68850181849e2fd1992a31dd20b7","text":"Describe the bug Currently the python 3.7 module (the other versions likely too, but I have not tested this yet) fails due to some numeric discrepancy caused by a recent upgrade of . This breaks ~26 downstream packages. I'm filing this to keep track of this for and am currently bisecting to confirm and test some other versions too. This is caused by a singular test failure: Upstream issue: Full hydra log: To Reproduce Steps to reproduce the behavior: on or . fails Expected behavior Tests should pass. Maintainer information:\nFixed by"}
{"_id":"q-en-nixpkgs-713f12e7f05284f893368b03375cb7fa2d05e9174d6884bc7904342486f867b5","text":", (nixos-unstable) Scanned versions: nixos-unstable: . May contain false positives."}
{"_id":"q-en-nixpkgs-86759565ae49f51d0052d48852268039038f6d4ce6a56356e1119a43fcb10efc","text":"libwebsockets 3.2.0 changed its behaviour in a way that breaks all versions of Mosquitto that use websockets, resulting in incoming connections being intermittent and very slow. It is fixed in their repository and will be in the next release presumably, but for now 3.2.0 is no good. If Mosquitto is the only other package using libwebsockets, I would suggest downgrading to 3.1.0, which behaves fine. Mosquitto 1.6.5, which should be released today, will refuse to compile with libwebsockets 3.2.0.\ncc\nmosquitto 1.6.5 fixes it - just haven't packaged it yet.\nMosquitto 1.6.5 does not fix this. There is a changelog entry which was written prior to this issue that is probably why you think it does have the fix. Sorry about that, I changed the changelog on the website post but not in the repo. The only fix for this is in libwebsockets itself.\nit's pretty awesome having upstream providing proactive updates like this - much appreciated!"}
{"_id":"q-en-nixpkgs-dd1fb7e44c49fd78a35543f6d037e65b7ec820b051c85e81399797f488ded154","text":"Describe the bug On opening Anki, I get this error My guess is the fix would be similar to (add ).\nIntermediate fix: I put this in my in :"}
{"_id":"q-en-nixpkgs-ad04fdf4e05e638a26c1ba74488d536cc5c6de804e8ee55e5396e66ecf8967b6","text":"I'm trying out NixOS 19.09 and notice that Enlightenment is still being built under X. Enlightenment 23.0 have some highlight. One of them being \"Massive improvements to Wayland support\".\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . 3. Ask on the on .\nIt would be nice to hear about the recent progress."}
{"_id":"q-en-nixpkgs-d2f5eda4f3528964ab47345a591d9b547d2d44b3b67c9e2a7d6f62dc87d98b17","text":"I am by no means an expert on database systems, so this may be a stupid question ... but I can't help wondering why does not refer to PostgreSQL version 11 on current versions of ? Is there a reason we haven't updated yet?\nMaybe or can answer this... :thinking:\nGood question. In my case I'm no longer using PostgreSQL 9 by myself and switched to 10 and 11 a long time ago, but I also wasn't aware of this because I'm not using the module from NixOS in my deployments.\nMade a PR for that... If you think it's too rush to include it to 19.09, then I can delay it to 20.03 As for , it really shouldn't be used. It should be split to , and , and / may be continuosly updated, but should be used only with explicit version. But this server/client split is not straightforward, so it was not yet done."}
{"_id":"q-en-nixpkgs-0e6c22a36e23f1b2cda19170877a73435ca7b585fdc6770b6757017664915ab8","text":"Describe the bug to the find call. This prevents things in the that are symlinks from being wrapped. Noticed this during office-hours when wasn't wrapped anymore. We should add some code like there is in or revert the . To Reproduce Steps to reproduce the behavior: a binary into or the thing and notice the binary isn't wrapped."}
{"_id":"q-en-nixpkgs-ac29c371cf3f92b0ca2a0578c68c52bfcf6673fcf4f0f17283c8c4726a7be4b9","text":"Describe the bug In , if you enable both xserver.services.desktopManager.plasma5 and , assertion fails. Failed assertions: cmt and libinput are incompatible, you cannot enable both (in services.xserver). To Reproduce Steps to reproduce the behavior: both and Expected behavior Failed assertion should happen Additional context Build with works fine. Metadata Maintainer information: kcalvinalvin a list of nixpkgs attributes affected by the problem a list of nixos modules affected by the problem\nFound the culprit. Can we disable by default so that users can choose? The default generated has the option by default so I don't see the reason why it needs to also be defaulted in"}
{"_id":"q-en-nixpkgs-5c1bda414073a19e7f0e09eb2e1ee956ab67423a1c1a72b1efbc237be352591b","text":"seems to be broken in more than one way. I get (also on ) Pinging maintainer:\nThat is, a build against current master with succeeds for me on NixOS 18.09 Will work on fixing that, but in the meantime you have a workaround.\nNot able to reproduce this one yet\nMh yeah I wrote a patch for to handle the case of not being present and after that, this error log showed up. Edit: I found my patch. Its really simple, just to fix the error. By building with that patch, I get the error log shown above.\ni build scylladb under redhat7(3.10.0-693.17.1.el7.x86_64) it fail with following information: i have enough disk for it, but i don't understand why it occupy so huge size.\nThe final build is for me. Haven't looked into why that is, yet. Could you open a separate issue for that? Thanks!"}
{"_id":"q-en-nixpkgs-a2de9c32192ae68c0c389e43e8337062a64416b0b540d31f135b6fd202bff088","text":"Describe the bug is broken on Darwin. git bisect shows (part of ) as the first failing commit To Reproduce Try to build on Darwin at <details<summarynix-build -A gobject-introspection </summary</detailsExpected behavior Able to build on Darwin Additional context <!-- There is a 10 years old bug report with the same error message --Metadata Maintainer information: cc:\nLooks like this upstream issue: . It's fixed on master so we should probably pick the patches.\nOur jobset needs to be building on the darwin platform if we don't want things breaking like this."}
{"_id":"q-en-nixpkgs-91620aab9c7889b98418c0353af14ece83171076467dfeaa2dee48fb58b9eacd","text":"Describe the bug Build fails when trying to install . I get: To Reproduce Expected behavior The package should install without error. Metadata Maintainer information:\ncc\nRelated to\nTried the following: It doesn't even bother about the C-core Version 6 that I asked: What am I doing wrong ?\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . 3. Ask on the on .\nRefreshed. My PR will fix this issue. Thanks for your patience."}
{"_id":"q-en-nixpkgs-130f3eed45275e57abee326ca0931c2dd96236e95e4be957a211eaf50a484a06","text":"Describe the bug Metadata Please run and paste the result. system: host os: multi-user?: sandbox: version: channels(hyper): channels(root): nixpkgs: Maintainer information:\nI've also encountered this. I bisected it to . says there's a newer version (1.) whose release notes say \"Fix two test failures on newer versions of gtk+/gnome-shell/mutter\". However, updating to it does not fix this test failure. :( With 1., the exact failure I'm getting is:\nWhen updating to GNOME 3.34 I did indeed notice this and tried the exact things did. We probably need to report it so it can be fixed, I just didn't have the time to do this. Guess the testsuite could be disabled.\nDisabling tests got it to build for me. I'll try to find some time to open a PR for that, and to report it upstream.\nReported upstream:\nThank you for your contributions. This has been automatically marked as stale because it has had no activity for 180 days. If this is still important to you, we ask that you leave a comment below. Your comment can be as simple as \"still important to me\". This lets people see that at least one person still cares about this. Someone will have to do this at most twice a year if there is no other activity. Here are suggestions that might help resolve this more quickly: for maintainers and people that previously touched the related code and @ mention them in a comment. on the . 3. Ask on the on ."}
{"_id":"q-en-nixpkgs-a72095b60a75988cfa3fa04c42a1d60f201a6cc12da854cbba470aa3b8dc7dfd","text":"Describe the bug sane-backends provides a configuration file, which is also produced by To Reproduce Steps to reproduce the behavior: Expected behavior No file conflict Additional context Add any other context about the problem here. Metadata Maintainer information:\nSomething isn't right there. I have in my , but I have no whatsoever:\nThe sane config is not written to /etc but only pointed at by the environment variable.\nThis is really just us being noisy about collisions, this one is harmless.\nEven if it's harmless, what would be the preferred way of getting rid of this warning?\nIt doesn't really serve any purpose for nixos users, so it was fine to remove."}
{"_id":"q-en-nixpkgs-2ddd34f7e7f363e7005e5d73d8119e5f4e0ad28df36ab3d487903781d3304061","text":"Describe the bug To Reproduce Metadata\nIn darwin building graalvm results in:\nWhich version of Nixpkgs is this? Please use the template...\nsorry I overlooked the metadata; updated it now;"}
{"_id":"q-en-nixpkgs-de9259ad6b82afb4a9e745baadc931c255a96aa505bad162c89b16006cb18492","text":"Running reports: To Reproduce on NixOS 19.09. . Metadata I ran but got .\nHello, I'm a bot and I thank you in the name of the community for opening this issue. To help our human contributors focus on the most-relevant reports, I check up on old issues to see if they're still relevant. This issue has had no activity for 180 days, and so I marked it as stale, but you can rest assured it will never be closed by a non-human. The community would appreciate your effort in checking if the issue is still valid. If it isn't, please close it. If the issue persists, and you'd like to remove the stale label, you simply need to leave a comment. Your comment can be as simple as \"still important to me\". If you'd like it to get more attention, you can ask for help by searching for maintainers and people that previously touched related code and @ mention them in a comment. You can use or on the relevant files to find them. Lastly, you can always ask for help at or at .\nI had this same problem today.\nI marked this as stale due to inactivity. &rarr;\nNot stale; this is still a problem on a flakes based NixOS system using"}
{"_id":"q-en-nixpkgs-e1fd5ed32324e3bf0d2e66ede9cced9b1a402db30cea851be149fd249b680c39","text":"Describe the bug As of 19.09, dolphin can't connect to smb network drives becasue kpasswdserver is unable to start up but chrashes kiod5. To Reproduce Steps to reproduce the behavior: dolphin in the address bar, or any other smb share in your network displays a timeout or access denied, check for the actual error Expected behavior Dolphin shows the smb share's content, using stored passwords if available. Screenshots : Additional context I guess this is caused by but I don't know enough about where kpasswdserver actually lives and comes from to fix this myself. Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information: cc\nThis does not only happen when trying to connect to samba shares, in my case this issue is triggered by the baloofileextractor indexer running in background, frequently causing kiod5 to dump core.\nThis is the backtrace of in my case:\nWe need to wrap in .\nWhat exactly needs to be wrapped there?"}
{"_id":"q-en-nixpkgs-54db289fe7a01363ea7d247b3e1b2d0f5b8d0eb718397e70b0b31338bd496954","text":"I've tried the following: which should assign an alias ( happily accepts the colon syntax in NixOS). Instead it tells me: \"invalid character ‘:’ in name ‘unit-network-addresses-lo:metadata.service'\" I also tried: which doesn't throw an error but also doesn't get me an alias. Is it not currently possible to alias the loopback interface via NixOS configuration, or am I missing something?\n(triage) any leads?\nMore than question I think it's a bug that it doesn't work. any thoughts?\nResolved in\nSorry, is this actually resolved? I try this in my config: And I get this: If I run the setup script by hand, it works fine. is this working for you now?\nuse instead: we cannot support arbitrary device names here due limitation of the script generator. Should be fine with though.\nI finally had a chance to follow up on this. Using the config that suggested above, I still get these errors: does actually get the address. However, boot-up and NixOps takes a lot longer, presumably because of the timeouts. So something is still not quite right here.\nThat device dependency should be not according to this source code: Are you running the latest NixOS 17.09?\nAhh, no, I’m running but from a few weeks ago. I’ll try again once I’ve updated.\nI still recommend to people to switch to since the scripted module is just a hack... well.\nSo with a recent , assigning an IP address to still causes timeouts. However, the IP address does successfully get assigned.\nif the service file assign ip addresses contains a dependency, something went wrong.\nI see the same:\nThe bug appears to be here: compared to this:\nTesting this patch:"}
{"_id":"q-en-nixpkgs-ea24d2ccb9af25c5e464b7958af9fcbb111090e6afcdeac3021a8e940995da2a","text":"Describe the bug Packages with potentially newer versions available inside Emacs will be overriden by ELPA packages. I am running an overlay with Emacs (27) built from master. On Emacs (27) master the package has the built-in version . The package provided by ELPA is . This is currently not an issue since Emacs has not been upgraded to 27. But when it does, this is going to cause a nasty \"issue\": To Reproduce Steps to reproduce the behavior: the load-path and notice that it includes the seq-2.20 elpa package. Expected behavior Perhaps not not included in the load-path... This is kind of the expected behavior, but I am raising this issue as from next version if not a new seq package is released to ELPA, people will end up in an infinite loop. Additional context Add any other context about the problem here. Metadata system: host os: multi-user?: sandbox: version: nixpkgs:\nAny ideas what to do about this case, is there any other packages where there potentially is a later version contained within Emacs and an older one available in ELPA?\nalready contains a few things mapping builtins that used to be separate packages to . This ensures that anything depending on will get the correct dependency graph while actually using the emacs builtins. The correct fix here is to map to , maybe with a conditional checking the emacs version.\nI've been using a locally patched nixpkgs to remove those seq dependencies, but the solution suggested here by is so much better that I went ahead and implemented it. Please can you also take a look at the PR to see if it resolves things for you? For me the breakage was pretty severe, so easy to notice (the package did not work; so I saw the problem at startup).\nThank you, I can verify that this is working for me as well. I am sorry to let more people experience this and not patching it earlier. I already spent some initial time finding the issue and then I made a patch within my own overlay like you. Glad to see it fixed, thanks :+1:\nI just discovered we have the same case for . Should that perhaps also be under a conditional? I guess if you are on an older version of Emacs you want to use the ELPA version, but if you are on the latest version you want the one bundled with Emacs as the version is more recent.\nCould you make a PR to nixpkgs?"}
{"_id":"q-en-nixpkgs-0fe64d36ae527e31229799765428d7b0f207ef632d96acd6fcdca79eeb342fe8","text":"Describe the bug Neovim crashes when resizing windows under some circumstances. To Reproduce There are many different ways to repro, these are some reliable steps to reproduce the behavior: to create a split to open a terminal to go to other window to maximize window Expected behavior Windows should resize and there should be no crash. Additional context Originally filed at but it seems to only be affecting NixOS. The crashes only started since I upgraded from NixOS 19.03 to 19.09. Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information: cc\nIf you have a known-good nixos revision and a known-bad nixos revision you could try the issue.\nThe earliest local generation on which this happens is from . The latest known-good generation is from . I'm not sure these are accurate though, I'd have expected this to have happened later.\nnarrowed it down to which was part of\nThat looks helpful. Maybe you can report upstream that the libvterm update is responsible, and ask if someone on non-nixos can reproduce it with that version of libvterm (or report which version they use).\nThanks for guiding me in the right path I opened a PR with a fix. :point_up:\nThanks for fixing this :)"}
{"_id":"q-en-nixpkgs-379abd98f16af88fac324b7dc10b395f09c588a6d359c4f42ac951e1b4ccaddf","text":"Project description A cross-platform SSH client with cloud data sync and more. Metadata homepage URL: source URL: license: unfree platforms: linux\nIt would be really cool, I need it too!"}
{"_id":"q-en-nixpkgs-2ecfc9c166620170b4f38e709f6c0538380b483c573bed35662f4510be799859","text":"I was getting errors with the version of dvisvgm bundled with texlive 2019 so I created a separate, texlive-independent version which works well. (If you're curious about the issue I was seeing it was related to compatibility between ghostscript and dvisvgm. I can be more specific if anyone wants, but reporting that issue is not the main purpose of this). Here's the build expression I'm using ( is ): How would people feel about a PR with a dvisvgm package uncoupled from the texlive version in addition to the one currently there? I could name it or or something like that. Note that I'm currently using this as a separate build expression but for a PR I'd put it in the texlive bin nix file. Note also that this package can be used identically to any other texlive package. The only difference to the end user is to get an option of a newer version.\nWhat is the issue, though?\nit's been reported . The issue is apparently fixed in dvisvgm 2.7.2 (ours is 2.6.3). Here's an additional MWE to test. The same example with v2.8.1 (in the build expression above) works fine.\nSo what you say, is that our dvisvgm is broken because of the incompatible ghostscript version. Then, I think, we should be fixing dvisvgm for all of our users, not just those who opt into using the separate package you propose. I would prefer to just patch the version from texlive, but, unfortunately, doesn't apply cleanly, so we would have to craft our own patch or pin to 2.8.1 (in this case, need to leave a TODO comment to remind to unpin for texlive 2020).\nTo clarify, the idea is to generate a patch from a newer texlive version of dvisvgm and apply that to the existing derivation (along with the code comment about this being a temporary fix until 2020)?\nDoes fix the issue for you?"}
{"_id":"q-en-nixpkgs-bfea1a7d9ccfec8c22e91e75dd8d0fa8ea7cb467e003566c7345c01ff13c9ece","text":", CVSSv3=9.8 (nixos-19.09, nixos-unstable) Scanned versions: nixos-19.09: ; nixos-unstable: . May contain false positives.\nUnstable/master: PR ; 19.09: (staging ATM, but should get into channel within days)."}
{"_id":"q-en-nixpkgs-e6397c9d6a5a7a78144ef893aea5dd6cf20624523aad3e66c8082542591a7084","text":"Describe the bug Fails to build with some not-supported instructions: <details</detailsTo Reproduce Steps to reproduce the behavior: (on an aarch64 machine) Expected behavior Kernel builds Maintainer information: May be caused by\nFixed already by bumping to 5.4.x, 19.09 channel just hasn't caught up yet.\nNever mind, I'm not very good at interpreting bisects.\nI had the same issue (I think?) when trying to install NixOS 19.09 on Raspberry Pi 3B+\nFixed in ."}
{"_id":"q-en-nixpkgs-4bd1c5cde3e53684c8a1a8b05a369a329905195593573ae569dad674947ad562","text":"Describe the bug NixOS manual combined fails to build since commit . This was introduced by the changes to the package awstats. This is the output: To Reproduce Steps to reproduce the behavior: Expected behavior It should be possible to build the nixos-manual-combined without errors. Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(terje): nixpkgs:\nthis issue was resolved in there is one more issue with docs - a warning What should it be? You can test manual builds with\nthis was my mistake :disappointed: I had the entry to the manual. Unfortunately I confused myself by checking the manual on another PR, and then thinking I had checked the manual for the PR. The issue was compounded by me being rather occupied on Christmas day (mostly with only cell phone but no computer), which I supposed in hindsight was a reason I shouldn't have merged the change on Christmas eve after the kids were in bed. I apologize for all the hassle. It looks like the option lost its during this PR... another thing I would have noticed if I had checked the manual... When is back they can create a PR to add a description along the lines of , unless this pressing enough you want me to file a PR today. Thanks again to all who helped out on this!\nno pressure!\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:"}
{"_id":"q-en-nixpkgs-a7982a2884c17f7ca384b5656e8ba8d463e63d692c4470b6c7427ea36ec9da91","text":"Project description A tiny IRC client Metadata homepage URL: source URL: license: mit platforms: linux"}
{"_id":"q-en-nixpkgs-c25c9fe1c9c0d8d4c4ba450a012eacb3da15e02949d569f1943141cfa4a43fdb","text":"Upstream , but we still have many packages depending on it. I know of no issues since the last release, so I'm not marking it as vulnerable yet, but the dependents would better start migrating in advance. Stable: it's unclear to me what we do about these in 19.09, but perhaps there won't be any notable vulnerabilities before 19.09 support ends :-) 1.0.2 is the default in 19.03, but we don't support that anymore.\nList of files:\nI think we should mark it vulnerable. Other EOL packages have been marked vulnerable because of the EOL before, e.g. Firefox, and if there’s any package we should be extra-careful with, its OpenSSL.\nI bumped it to fix the last CVE minutes before opening this ticket, so I was obviously biased against marking it immediately ;-)\nBTW, it's still a decision to make, as there's surely some LTS distro that will keep shipping 1.0.2 for some time, so patching the worst probably won't be too difficult. I'm personally not willing to invest significant time into it, though :-)\nBest way to motivate somebody to volunteer is to mark it as insecure. ;)\nOK, I think the we've given enough time and now we should press on a bit:"}
{"_id":"q-en-nixpkgs-c5c997d3535e4df6c3c7228126b65bd0c086c334db7db18673ee7d225aac0cd3","text":"Describe the bug Not able to install pycuda. Error To Reproduce Install pycuda Expected behavior install pycuda with dependencies Metadata Please run and paste the result.\nIs nobody having this issue it seems is not solved :(\nI'm having this issue, incoming PR. Likely half of the problem is that the cudatoolkit is pinned to 7.5 and cuda root is not passed during the build process, which is not compatible with newer cards. The other half of the problem is that python3 boost isn't being linked. I see boost is mentioned as not being necessary for pycuda in the docs, so I'm wondering if we can just remove it.\ntry and let me know if it works for you"}
{"_id":"q-en-nixpkgs-25b72ea275e0d3bbbef286ad13ed4dfb2753bc81a87cef24c01e9a02e1ddeab7","text":"Project description Desktop client for the LBRY protocol. A YouTube alternative that uses the blockchain. Metadata homepage URL: source URL: license: MIT platforms: Linux\nHello, I'm a bot and I thank you in the name of the community for opening this issue. To help our human contributors focus on the most-relevant reports, I check up on old issues to see if they're still relevant. This issue has had no activity for 180 days, and so I marked it as stale, but you can rest assured it will never be closed by a non-human. The community would appreciate your effort in checking if the issue is still valid. If it isn't, please close it. If the issue persists, and you'd like to remove the stale label, you simply need to leave a comment. Your comment can be as simple as \"still important to me\". If you'd like it to get more attention, you can ask for help by searching for maintainers and people that previously touched related code and @ mention them in a comment. You can use or on the relevant files to find them. Lastly, you can always ask for help at or at .\nStill not packaged.\nWork has been started in .\nI marked this as stale due to inactivity. &rarr;\nstill actual\nPR\nPR closed, anyone else willing, reopen under a new PR and try your luck at the merge roulette\nSmall update, I'm currently working on packaging LBRY from source (it actually seems possible).\nOk, while not from source (that didn't work out) LBRY has been packaged."}
{"_id":"q-en-nixpkgs-32b9c12cf236663fa5e52d69599185aa85fc32dde23fc90efc68f1cb5e2610af","text":"Describe the bug NixOS’ Reaper package is not configured to work with Jack2 backend. To Reproduce Expected behavior Reaper should manage jack backend. See for instance . For now, Reaper only says in Reaper Preferences/Audio/Device. Additional context Everything is working fine on my system with jack2 (Ardour, Supercollider, VLC, etc.), except Reaper. Metadata"}
{"_id":"q-en-nixpkgs-7a28f1cdefeadd38c0bebb7e32b3ff5599474571c5e5b42680dedb3176fd2f34","text":"Describe the bug When I launch using a keyboard shortcut, it didn't get any context, resulting in it displaying strangely: ! If I launch it using the menu it displays correctly. For other applications, depending on what they requires, they sometime just don't open. EDIT : I have the feeling that the reason that is not opening when using the keybinding is the same reason. To Reproduce Steps to reproduce the behavior: a shortcut to Konsole konsole with the shortcut Additional information While starting with the shortcut, this inputs is visible in : Expected behavior The program should behave the same way as when launched using menu. Metadata Maintainer information:\nWoah, it doesn't seem to even have a PATH environment.\nNew way to reproduce the issue : when the feature to restore previous window after a logout/login is active, it restarts the application with the same issue.\nApplying the fix from fixed this issue.\nI cherry-picked and it fixed this issue for me; however my KDE windows are all still borderless.\nHmm, that sounds like kwin died."}
{"_id":"q-en-nixpkgs-282db782783a6f2783348dea1e2ac8df8bfbf6e4347693b81e414bd68d48ec11","text":"Describe the bug bazel broken on darwin in nixpkgs HEAD () To Reproduce Steps to reproduce the behavior: to build bazel on darwin Expected behavior Able to build bazel on darwin Additional context might be related to Metadata Please run and paste the result. Maintainer information: cc:\nThe bazel build is passing LDFLAGS flags to the compiler instead of the linker. Those should be prefixed with or moved somewhere else.\nClang used as a frontend should accept linker flags.\nFlags were in ."}
{"_id":"q-en-nixpkgs-0fbf4420b3d327a5e291141f614dcf34f46da062b287895edec61c56913c4d6d","text":"Describe the bug Installing nomachine-client fails To Reproduce Steps to reproduce the behavior: to environment.systemPackages Expected behavior It should build the derivation Screenshots If applicable, add screenshots to help explain your problem. Additional context Add any other context about the problem here. Metadata Please run and paste the result. Maintainer information:\nI tried overriding the hash (any tip on how to more properly do this would be appreciated), but still got build errors in my under environment.systemPackages result:\nThere's a PR to upgrade : I this to my and it seems to be working for me:\nYep, this is because NoMachine removes each old release as soon as they make a new one. To combat this, I'll add to the sources list when the update is done.\nI was finding at the time that even overriding with the latest URL and hash from the nomachine website, that it was failing to download the whole blob. I could do it from my browser, but it seemed to abort/not download the full size via nixos\nOh, that's strange.. Thankfully, it seems to work just fine now :)"}
{"_id":"q-en-nixpkgs-3eb3eb4c4d2abe3d01b7324388155d0f073c415b5b9dd16133147a34558d3cdc","text":"Describe the bug I've tried building two packages from source on my new MacBook Air running macOS 10.15/Catalina ( and a custom expression for ), and both times I get the same set of errors involving Xcode/iOS simulator/permissions/etc. It never seems to start to the build. Example: I don't know for sure whether it has to do with Catalina, but I never received these errors when using 10.14/Mojave. To Reproduce Steps to reproduce the behavior: Nix on a Mac running Catalina. I used the instructions . to install a package that has a source-based build for Darwin such as . Expected behavior The package should install or at least give me errors related to the build rather than Xcode/iOS simulator/permissions/etc. errors. Metadata Maintainer information:\niterm2 is currently broken on Catalina. I remember to have it fixed locally months ago but didn't find the time to open a pull request.\nI was just using iterm2 as an example. The errors I'm receiving are generic and occur for any source build using Xcode on my new computer running Catalina. Separately, I also remember iterm2 being broken while I was using Mojave. These are different errors.\nCould you test my pull request that should fix this issue (at least for iterm2)?\njust tested and it works, thanks! But what are and ? I can't find those documented anywhere."}
{"_id":"q-en-nixpkgs-9857396b27856000ce36d10cfafe61e5c6fe2bf9bbfd96b34dd6e3b10b2c7018","text":"Update certbot Current packaged version is 0.39.0, upstream version is 1.2.0.\nFix in"}
{"_id":"q-en-nixpkgs-ca36f370019c3fe9e3914464896a773d6736b71d7a2f1828214ec8a512527e5c","text":"I'm thinking of bringing this line back: and getting rid of nixos/modules/misc/extra- Does anyone have objections? I fully buy the ability to pass arguments in the module system using , but if it doesn't work (and breaks in a horrible way) for the most common use case ( in imports), that seems bad and the cost of restoring the old way seems pretty low. Anyone have objections? I can make the change if nobody objects. Ideally, I'd like it in 15.06, since it's getting in the way of some work I'm trying to squeeze into 15.06. cc\nIMO our only choices are to do this or to set as a prerequisite (perhaps enforced using ) that points there. No strong preference either way.\nYeah, this is probably the best solution.\nCool! Do any of you know of VM tests that would suitably make sure I don't break anything too obvious by making that change? I'd obviously rather not make the change and then discover a week from now that I broke someone's key use case :smile: I'll be making the change this evening at this rate.\nAny thoughts on having set using scopedImport? That way stays NixOS-independent. But not a huge difference IMO.\nI don't see how adding back could break anything.\nFamous last words! :grinning: But I agree, as long as it evaluates at all it's almost certainly fine.\nThat sounds more complicated than the one-liner suggested by"}
{"_id":"q-en-nixpkgs-f2f9f290997cf6f09837bf2fc15b4592160296b323934893061ec10258581bff","text":"Describe the bug The description of is: Path to the unix socket file on which to accept FastCGI requests. This option is read-only and managed by NixOS. So i can't set it. But how can i figure out how the socket path is now? Please add an example to the description! To Reproduce Steps to reproduce the behavior: option desciption Expected behavior Useful description. Metadata same in 19.09 and 20.03 Maintainer information:\nThe point of the option is that you don't need to find out, just reference it. That being said I agree with your issue and it would be helpful to set an example\nThe point of the option is that you don't need to find out, just reference it. Sure, that works in but i have that path explicitely in my webserver config which is next to the content, e.g. . For reference: The path is actually as i set it before: So i can just remove the option from my ."}
{"_id":"q-en-nixpkgs-03d61552963971a8d00edd65cb2def5fa30ddd55b3cff600351a98245bed9e2d","text":"Describe the bug Trying to build a combined TexLive that includes Biber (e.g. ) results in failure on darwin. This seems to be the relevant error: To Reproduce Steps to reproduce the behavior: , where Expected behavior The build should succeed. Additional context The build succeeds in the example above if is omitted. Other schemes that include Biber, like , fail with the same error. Metadata system: host os: multi-user?: sandbox: version: channels(matt): nixpkgs: Maintainer information:"}
{"_id":"q-en-nixpkgs-5862abaa7de879f8fb0bbccfab049f1933d8b74f8adb69336a0fe327bce78809","text":"Describe the bug Haskell derivations: behaviour of in regards to changed in 20.03. Projects relying on the old behaviour will break. A number of Haskell projects I work on use to make test dependencies and build tools available in the environment when invoking . PR changed the implementation of the attribute which subtly changes the effect has. Using instead is a workaround for this (or possibly the function I should've been using all along). I believe I was using incorrectly but it is difficult to be sure, because the expected usage of and is not documented (AFAIK). Using In 19.09, ( ++ ) produces: In 20.03, ( ++ ) produces: is present in 19.09 but not 20.03. Interestingly, is not present in either. If we instead use : we get the following results in 19.09: and in 20.03: and are present in both. So, using in and does exactly what I want (with the version being a bit cleaner). The \"fix\" for this issue might just be to document the intended usage of: , , , , , , , and a bit more clearly (if not already - sorry, I couldn't find any). To Reproduce Steps to reproduce the behavior: (). it and cd into the directory. the following commands in a : where the invocation is one of: number of convenience make targets are available too. Expected behavior I expected to make and available in my shell.\ncc as the author of , (who I think wrote the original version), and (who did an extensive review) We should probably have documentation for this, or at the very least a note in the release notes for 20.03.\nSide question: how did you print the result of in the context of your actual package?\nSo I created a small nix file something like: Then:\nThis issue has been mentioned on NixOS Discourse. There might be relevant details there:\nI think this looks correct to me. Nevermind the build-time vs run-time distinctions I am usually harping about. The most important thing to understand for this is just like the regular generic builder always did, the now respects the different type of inputs parameters stating Haskell vs non-Haskell, and uses that instead of the predicate it used to sift between the combination of everything before. tl;dr and items always don't become in env, and instead get passed on to the , while and do become . On the build-time deps side, including the tools, whether or not something was written in Haskell doesn't matter, so they all become ."}
{"_id":"q-en-nixpkgs-40dcb6d317f33e0a6f2c1784cc6759514423c34cb8a1fb9083abc68dfa212c28","text":"Project description A modification of the Mesa Vulkan overlay. Including GUI improvements, temperature reporting, and logging capabilities. Metadata homepage URL: source URL: license: gpl3 platforms: linux\nI tried myself at this and have a working package now for NUR based on But I haven't yet figured out how to create a 32 and 64-bit version. Depending on the game both are necessary. To just get a 32bit version I currently just use: instead of just\nIt might make sense to just build the 32bit and 64bit packages separately and then merge them together in a wrapper.\nokay I did something similar, though not with a wrapper which is probably a little hackish I am not sure: but this allows me now to use mangohud with 64 and 32 bit applications\nThat seems like a reasonable solution, although it is a little bit hackish. I'm running into a similar issue with , and it's not really clear what the best approach is. recommended keeping the two builds separate for vkBasalt. However, to ensure we can install both builds we would have to do extra patching, and use separate names for the 32bit and 64bit build. Users might also expect to have both builds in one package like on other distros. I'm leaning towards something like this or like what I've done in my .\nwas just released but that changed quite a few things around. instead of using two different binaries it now uses  to decide where the 32/64bit libaries are, but it seems to me that this always expands to 'lib' on NixOS This is what I have so far. Having to ignore collisions is also not really that cool. Well having a running 0.4.1 version would be nice enough, just thought I write about this somewhere\nThe commit in question:\nI marked this as stale due to inactivity. &rarr;\nStill important to me."}
{"_id":"q-en-nixpkgs-15e3d21ee90972451633c5cd03384a60a8f16deeb42aa2f933e07f1f7699b889","text":"Describe the bug When writing , it was important to ensure that the .lego state directory was common to all certs to ensure that new accounts were not being created for every certificate. The rate limit on new accounts is 10 every 3 hours. This is fine for most cases, but in my deployment I have 30 different domains to request certificates for, and that's after a lot of deduplication and grouping of domains. As of this has been changed such that all certificates get their own state directories, and thus their own account credentials. As pointed out, the account directory can be symlinked using preStart scripts if necessary. I feel this isn't a matter of migrating user data but more about maintaining compatibility with existing setups like my own and also avoiding rate limits in LE's service. This should be part of the module's configuration by default.\ncc and whoever else I'm missing...\nI have been focused on getting a mail stack in Nixos working over the past week and I missed that PR and other things :/ sorry about that. I would've commented if I had seen it sooner.\nSeems like reverting for now would probably be for the best; multiple certs for the same domain seems like a less important feature than not spamming LE with account creations. It should be possible to reuse the same account while keeping the certificates split, but maybe we'll have to make some ugly symlink forests? I wish lego's CLI was more explicit.\nWell that or, we can patch in symlinking of accounts directories back a folder. I think it wouldn't be too hard, I can prepare a PR now\nYou before self-merging the PR and backporting it to 20.03; could you please subject your changes to critical infrastructure to wider review in future? It's quite frustrating to work with Let's Encrypt engineers to make the NixOS service well-behaved, only to have a change merged and backported that runs roughshod over rate limits and can potentially result in hundreds of certificates being renewed at once on upgrade just to avoid incorporating , with less than three hours between the PR being opened and merged, and only one recent contributor to the service being pinged. It's a lot easier to fix issues before merge than it is to rush to fight fires after something's hit the channels...\n: merging every certificates in a single directory leads to this issue: It was considered blocking for 20.03 because it broke a behavior that was authorized and expected in simp_le’s age. (The idea of symlinking the account folder was proposed, but since noone reacted noone implemented it I guess)\nI saw that issue and I don't intend to reintroduce it, but sharing just the accounts folder would fix this issue without regression. Of course, I will test and ensure that is the case on my system first but rest assured I have no intention to completely undo\nThanks :bow:\nI created a I still don't fully understand the use case behind having multiple certs for the same domain? Surely, on a single system, you would only want one certificate per domain, even if they are for different services? Also I did think of an issue with the .lego folder as a whole... if a user uses different user and group for 2 certificates then 1 of them will own the directory and the other will not be able to write to it. I don't know if we want to support that sort of configuration? It would certainly prevent us optimising around rate limits, but it is solvable if the .lego directory is under\nYou can use different key types (to serve both RSA and ECDSA certificates for the same domain), or perhaps set other alternate names. In terms of directory sharing, I increasingly think that the only permissions option that makes sense from a privilege separation point of view is to have unconditionally owned by and just selectively add other services like nginx to the acme group for read access (per ).\nYeah I had read that earlier and I agree that an user + group would make sense. How do people feel about a PR for a complete revamp of the module? This was suggested back when I finished DNS-01 due to the number of changes and the complexity of the module at this stage. I would be tempted to at least try and do this (now that I have the time again), and see what can be simplified. It's becoming quite the monster.\n: if you want to completely rewrite the module, you might be interested in my recently listing of some issues regarding ACME: I think the last one might be very annoying: once a certificate is created, changing the list of domains in the certificate won’t work (in other words, in the current state of the module you cannot add or delete domains from a certificate after it’s generated)\nYep also started tracking that issue already :) I think that one is resolvable by writing some file that hashes the domains + extraDomains, and takes action accordingly, but RemainAfterExit is a bigger bummer. I'll continue conversations about a rewrite over in that issue.\nIt was reviewed by and Catching ratelimits is not something one can easily test or fix. If one has 30 certificates they will run into certificate creation limits, in which only SAN certificates would help. The needed migration code was posted on Discourse and the Github issue for those that run unstable (Lego in ACME was never part of any stable release).\nPLEASE backport this to 20.03, since it's a serious bug, with effectively completely breaks the lets-encrypt-certificate-renewal in 20.03 if more than 10 certificates are used on a server.\nIt was backported already. Are you actually experiencing issues? If so; please open a new bug for that\nI've updated a server to 20.03 today, and I'm experiencing exactly this issue."}
{"_id":"q-en-nixpkgs-05fe84689d6e3fedc34153d44e15966746c2057649c84028cd15a23b4059ffed","text":"Describe the bug Systemd ships dbus system services (in ), which are used for dbus system activation (see ) This includes some of the units shipped with systemd itself: We currently don't have the proper tooling ready to make these available in NixOS. Also, we shouldn't blindly enable all. For example, we might not want to enable `share/dbus-1/system-services/org.freedesktop.resolve1.service, if resolved is enabled. Originally discovered in , but it always hasn't worked, so it's not a regression. Notify maintainers\nI wonder if that's why every program I have that uses dBus in some way or another, spits from time to time:\nCan you tell me which programs those are?\nSure, from , prints: I'm using as my display manager and I'm using as my desktop environment but I don't use . Not sure if that matters.. EDIT: Currently I can't identify another program that spits this message, although IIRC I saw it coming out of other programs as well. This could be an upstream issue as well.\nIt does seem that somehow picks up the systemd configuration; though I do not fully understand how it's just missing the when is enabled\nAh apparently all applications are loaded into dbus by default? Not sure how I feel about that ... so the only reason why resolved wasn't working is because we didnt included unit file in Easy fix! All the other services already work!\nIt seems we do, but we shouldn't do that. If we try to not restart , changes in there don't get propagated. It should just use nscd like everything else. can you file a PR removing this?\nreopened, as got reverted.\nI've found another example for: prints:\nAnother symptom is:\nI'm not sure if that is a symptom of what I am solving here. I recall recently running into the same issue and I think it's a problem with how we set up user environment in certain desktop managers.\nShould I open a distinct issue then?"}
{"_id":"q-en-nixpkgs-049f5518a8b3ff5b1480bf558f124ff5cdf1f5eeaf205e7017f56df9333e3ca7","text":"Describe the bug There is no AWS AMI for NixOS 20.03 To Reproduce Steps to reproduce the behavior: Check or Try find it in AWS AMI console Expected behavior 03 is created and uploaded to AWS\nCC I think knows how to do this part.\nYep, I have no idea how to do this for NixOS, so I'd need someone in the know with AWS AMI for 20.03. This actually was one of my concerns for 20.03.\nI know how to build the release AMI and put it on a personal AWS S3 bucket (I'm using one now for my NixOps AWS cluster), but do not have the permissions to upload it on the official account. about 2 months ago I followed this process, which you can use if you're looking to unblock yourself: then follow these instructions to upload using : Then on NixOps you set to your custom AMI. You can also use this to bake in pkgs to the AMI so there's less startup time if you know you'll be downloading specific things on provisioning startup otherwise."}
{"_id":"q-en-nixpkgs-a604952ec6bdcbf36ee446f1ebcd895d04c697b09de34117dd10270bba4501db","text":"Describe the bug A clear and concise description of what the bug is. The package fails to build. It errors out during the installation of bash completions with \"Permissions denied\". To Reproduce Steps to reproduce the behavior: -build -A mosh Expected behavior Successful build Notify maintainers Metadata on Maintainer information:\nThe breaking commit is the bash-completion update from 2.9 to 2.10: . Any idea how to resolve this?\nI think you meant to link . Seems to be a regression from incorrectly dropping by ( in ). The \"wrong\" prefix comes from . A possible alternative would be to drop the dependency on for (if there's a correct fallback - since we don't actually the prefix from anyway).\nI think removing this patch was correct, as the functionality introduced in that patch was to 2.10, however a program needs to define instead of when reading variable, which may necessitate patches (which could be upstreamed). It's unfortunate that the upstream decided on different approach than what was done previously, but trying to preserve the old behaviour would cause more issues in long-term. Let's not overcomplicate things when some of upstream projects already have code that uses when using = 2.10. I imagine this is the line that needs changing:\nsorry, my assumption about the patch was wrong then :o And thanks for the fix :) Edit: And I agree that we should follow upstream then."}
{"_id":"q-en-nixpkgs-8c238ed86bbcf3d7d0d650955b8d9f67d59ee0c10934294ce398a088bacc67cf","text":"Describe the bug do not run. To Reproduce Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata Please run and paste the result. Maintainer information:\nmeets the same error.\nI think I might have the same issue with solargraph: The same command from nixpkgs-19.09 seems fine: Possibly due to the bundler changes in ? EDIT: Hm. But in a docker container (lnl7/nix:2020-03-07), runs fine. Something in my environment? Or darwin weirdness?\nI had a similar issue and I found that my global bundler config was the problem. changed the generated binstubs so that the environment variable is no longer set which I think means that bundler will now use from the global config if it exists. This seems to result in bundler looking for gems in the wrong location. A possible workaround is to set using an environment variable instead of using the global config since this value will be unset by the generated binstubs but will still have an effect when bundler is used from somewhere else. There is also the environment variable which causes bundler to ignore all configuration but that might have other unwanted effects.\nI marked this as stale due to inactivity. &rarr;\nDo you still experience the issue? nixos tests of the service are passing and I'm unable to replicate on the current nixpkgs or even in commit\nYes. and don't run. I'm in nixos-unstable .\nI removed in then both programs run fine.\nHopefully or a similar change will get merged and prevent these problems. I've seen a few issues like this get reported."}
{"_id":"q-en-nixpkgs-b8dfb9ac7de21a8ef78d5da32dbd0c0bc94f6df732f9262930ed7fc75efe37a6","text":"Hi, i need to use open-zwave as a dependency for a build, but the version built in the Nix package is a specific fork for supporting home-assistant, which does not work for my package build due to linking errors. I will happily contribute a new package, but I'm not sure how to proceed, with regards to how it would fit in with this package, or if indeed the best way would be to modify this package to use the upstream repo. Pinging people who have maintained the package, apologies for the noise!\nHello, I'm a bot and I thank you in the name of the community for opening this issue. To help our human contributors focus on the most-relevant reports, I check up on old issues to see if they're still relevant. This issue has had no activity for 180 days, and so I marked it as stale, but you can rest assured it will never be closed by a non-human. The community would appreciate your effort in checking if the issue is still valid. If it isn't, please close it. If the issue persists, and you'd like to remove the stale label, you simply need to leave a comment. Your comment can be as simple as \"still important to me\". If you'd like it to get more attention, you can ask for help by searching for maintainers and people that previously touched related code and @ mention them in a comment. You can use or on the relevant files to find them. Lastly, you can always ask for help at or at .\nWe should replace the package with an official version and make use the fork."}
{"_id":"q-en-nixpkgs-94f9e510faa2ec0376f1ee1d4a93c7f58b6b886dc79e5999058e031db5a16858","text":"Describe the bug KDE has trouble with some SVGs, particularly the one bundled in alacritty. To Reproduce Steps to reproduce the behavior: alacritty Expected behavior The icon was properly rendered. Screenshots See picture of gwenview vs. inkscape vs icon in my taskbar. ! Additional context It seems alacritty bundles a png for KDE, as well as a high compat svg, I'm currently trying to make an overlay to fix this. See:"}
{"_id":"q-en-nixpkgs-c66150909c53f7e65b10208f8f7d34f284bef9a9b5c0d2565d117dd4bf494257","text":"Describe the bug Since switching from to I needed to switch from to , otherwise the system does not boot anymore. To Reproduce Steps to reproduce the behavior: system on lvm-inside-luks using either or channel password - gets accepted correctly - LUKS seems to be working does not boot - complains about missing root device Expected behavior Would boot :) Screenshots : ! : ! Additional context All my config for this machine starts from - I link this file in place of Notify maintainers maybe ? <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata Please run and paste the result. ^ using channel with kernel right now Happy to help debug this if needed :)\nWhat filesystem are you using? If it's a mainlined filesystem, I don't believe hardened should've broken that. But it's probably depending on which is disabled in hardened.\nThanks so much for the reponse :) It's for , hopefully still mainlined ? ;) I can try to compile the kernel with set but otherwise with hardened settings enabled and see if it helps ?\nSorry, this is my fault! Yeah, my guess is that we need to re-enable AIO. I did notice some additional harmless-seeming warnings at boot that seemed AIO-related when writing (which switched to a patchset that disables AIO by default), but I guess it didn't affect any of my systems because they use ZFS. I was hoping AIO was only required by marginal things (like Oracle...), but LVM is pretty important. Given that nginx also uses AIO per , it might be worth explicitly turning it on in the default shared Kconfig (as I did for System V IPC in the original PR). I'm not sure if the AIO code is known to be particularly crufty by Linux standards or if it's just a general attack surface reduction thing; a few of the linux-hardened defaults are a bit better-suited to the Android environment than a desktop Linux distro. credits to the original linux-hardened author's Android-based which makes me suspect this might be the case here. Here's a nixpkgs diff to re-enable AIO: It'd definitely be good to know if this fixes things for you. should be enough to test this. If you don't want to use a custom nixpkgs then you can do a fiddly thing like the definition of  in .\nAlso, this would be pretty easy to test for, so if this change fixes it then we should also update the hardened test to use LVM.\nHeya, good news ! One kernel compilation later - I'm back to running thanks to you :) I tried the route and indeed it worked. For the test - I think that would be awesome, to catch things like this in the future :)"}
{"_id":"q-en-nixpkgs-2c2d3c67e67aad98443cb70405a6092255720334db23ce7cafa43bd44be3e38c","text":"Describe the bug Riot-Desktop -Security & Privacy Under the \"Message search\" heading, observe: Riot is missing some components required for securely caching encrypted messages locally. If you'd like to experiment with this feature, build a custom Riot Desktop with search components . it links to here: Notify maintainers cc: Metadata Maintainer information:\nWould be fantastic to have built with . I tried building it myself but couldn't do it. :cry:\nJust stumbled across this myself. What's the current state on this? What needs to be done before this can be fixed?\nSo, element seems to use its own build system to build these native dependencies. It tries to download and build those itself with cargo and stuff. Seems quite complicated to package.\nI would like to add that support for desktop notifications would be also appreciated. (They showed up on Arch's version, at least) Edit: This problem has been resolved since\nI made some progress. This solves the issue of downloading the hakDependencies. Unfortunately they are not locked in any way, so we need to lock them at packaging-time (is that even a term?) by adding them to and Since yarn2nix always runs yarn with , we can safely ignore the upstream reasoning for fetching them outside of yarn dependencies. I'm now at the point where it tries to build seshat in the element-desktop derivation, but unfortunately they don't provide a Cargo lockfile. I'm working on resolving this problem upstream: <details</details\nIt turns out to be much more complicated, because the cargo build of seshat-node-native calls , so we have multiple layers of npm calling yarn calling cargo calling npm. With lots of things being downloaded and executed. It's a huge shitshow.\nIs there anything we could do to help fix building with encrypted chat search working on NixOS?\nI poured some (many) more hours into this today. I can build the two native modules, matrix-seshat and keytar, both in a seperate derivation and as part of the element-desktop build. However, I can not build them correctly against electron's node lib, so when starting the compiled thing, it refuses to load the module. Attempting to build against electron gives a mystic error. Or the second approach, with everything in one derivation using the hak tool: <details</details<rantHonestly it's just a horrible build system. hak and neon and so on don't give any consideration to users who want to build the libraries in an environment without an internet connection. </rant If you can find a fix for the error, there shouldn't be much work left.\nI think I found the solution to that error: Upstream is using Electron 12.x while we are using Electron 13.x.\n! PR incoming :sparkles:"}
{"_id":"q-en-nixpkgs-d065b273e5b70fd9c37d6e3ee2c0244e52c250e0e26a8c01fc82d05a26124078","text":", CVSSv3=9.8 (nixos-20.03) Scanned versions: nixos-20.03: . May contain false positives.\nmerged - closing"}
{"_id":"q-en-nixpkgs-e907fca76682191aa7386debae16fe43ae8f8f6739072add8a963d44a52b0f3a","text":"Describe the bug The package is supposed to vend a binary called . Instead it's vending a binary called . The upstream repo has a Makefile that explicitly passes to . I don't see any equivalent in our package (which bypasses the Makefile). This is a problem both because of breaking expectations, and also because the package makes the assumption that provides a binary called . In particular, the script tries to invoke the binary as part of the subcommand. To Reproduce Steps to reproduce the behavior: what binaries it provides Expected behavior It should provide . Notify maintainers Also the maintainers, in case the resolution is to change to use the name instead of : Metadata system: host os: multi-user?: sandbox: version: channels(lilyball): nixpkgs: Maintainer information:\nIt's possible to override the buildPhase in to run make. However in this case it might be easier to just perform the remame with a in ."}
{"_id":"q-en-nixpkgs-91cb2987063646378ab9585038192850b040f03025e4c5f6381607fb97e2e2e0","text":"Describe the bug Error build php packages. To Reproduce Steps to reproduce the behavior: Run Notify maintainers Build error occurs after this PR Build log:\nAfter this PR error build php. cc\nfixes it"}
{"_id":"q-en-nixpkgs-932cfef9654d7a51e3161e6cfd2555c3f970aca01625e5a622261ca144671872","text":"The I wrote is failing on (): The problem is that when machines reboot in quick succession, consensus (and cluster leadership) is lost permanently, and 2-of-3 servers cannot continue to serve requests. I have debugged it and found that it is already broken since (), but it was noticed only now. :red_circle: This is why we should run the relevant tests for every PR and should show big red warning signs when they fail! In any case, I have bisected the problem to be commit , introduced between Consul and . I have filed an upstream bug: I also found a way to make the test pass: But I don't think that's good, because I think this is a fundamental issue with Consul that affects operations (the 2-of-3 HA promise being broken), so I think it should be fundamentally fixed, and that the test is right. NixOS 20.03 still has Consul so it does not suffer from this bug, but the next NixOS release will. CC as previous consule package contributors.\nIMO we shouldn't even use attr, but always refer to concrete version, like or . Seamless major consul upgrade is fine in docs, but disaster in reality (happened twice to me). That doesn't quite solve this particular problem, I know.\nYes, the key thing is that Consul doesn't permit downgrades of its state in . I would thus also always recommend to upgrade non-point-releases separately from NixOS upgrades.\nHonestly I kinda wanna make a few changes to most of the Hashi packages and tests soon to provide more stable versioning (and reliable testing). I'll start cleaning it up this weekend hopefully\nUpstream has a great explanation and recommended fix: Test fix and more tests in PR . Key change: . I recommend every Consul user to read that linked comment and the updated test to understand the upstream-recommended rolling reboot/upgrade procedure."}
{"_id":"q-en-nixpkgs-66dffd12d4da0608835df21130726191e7fafd0fa2a49f730574eee8bef5d3b1","text":"Describe the bug It should be the path to ca- To Reproduce Expected behavior No error happens. Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(mnacamura): nixpkgs:"}
{"_id":"q-en-nixpkgs-2e6971d0083b09c096aa32dc9e54777743a9b1f9a68972064269bff531ec05e7","text":", CVSSv3=9.8 (nixos-20.03, nixos-unstable) Scanned versions: nixos-20.03: ; nixos-unstable: . May contain false positives.\nmerged\nNeed to keep open for 20.09\nMerged to 20.09 too. Closing."}
{"_id":"q-en-nixpkgs-d91c6c16ca00286bc892396f27f7fdc19ad456b9d3e1b41acb45ec2398889f42","text":", CVSSv3=6.1 (nixos-unstable) Scanned versions: nixos-unstable: . May contain false positives. Cc Cc"}
{"_id":"q-en-nixpkgs-dcd05c07d48f2cecdb762d0b04265970b043b1be404307e37663b7096800488a","text":"I am failing to install on right now: What more details should I provide?\nperf fails on kernel 4.1 and 4.2 only\nfor an (arguably hacky) fix."}
{"_id":"q-en-nixpkgs-95556984319c083effb7dd4e1f2e86ca5705f6c43099cc9c5b7dd7c9da49f523","text":"Project description Netdata agent plugin Metadata homepage URL: source URL: license: GPL 3.0 platforms: unix, linux, darwin We already do have netdata, however it seems to be built without support for the cloud agent that allows it to stream to the cloud or other netdata servers.\nI have a local checkout with some edits that allow me to build netdata with cloud support. Unfortunately they also require a change in the libwebsockets build expression. Should I open a PR to integrate my changes into nixpkgs?\nA PR would be great! I'm not a maintainer, but if they decide it's out of scope for nixpkgs, perhaps one on NUR might work - I'd try for a PR here first though\nGreat, I think having a configuration option that is per default would make sense. Maybe ?\nYeah, that seems to fall in line with the various flags present in other packages\nI marked this as stale due to inactivity. &rarr;\nUpstream has incorporated some of the suggestions in this thread, I intend to pick up on it again soon\nwhat would be needed to make this work? Perhaps I can help?\nThe newest major release of Netdata (1.30.0) includes a cloud link implementation without the need for vendored dependencies. Unfortunately, it is not currently in a reliable state because of a bug in the custom HTTPS implementation. The has already been merged into Netdata master and will be included in the next release. I plan to make a PR that enables cloud support as soon as that release is available.\nv1.31.0 dropped 12 minutes ago :) Anything I can do to help?\nThanks, I've already gotten a notification for that. Planning to make a PR when I get off work today. When my PR is done, it would be helpful if you could test my changes on your machines.\nJust FYI, I have opened the PR that enables cloud support a few days ago. If you are still interested it would be greatly appreciated if you could test if everything works for you and maybe drop a comment.\nSorry, missed this but the pr looks good! I propose we backport this to stable?"}
{"_id":"q-en-nixpkgs-d74e2392d010cee873b6b597603a8540dd54cfd0606091c2145850a91d8ac664","text":"Describe the bug , a package which requires PySide2, fails to install, either system-wide, nix-env local, nix-env global, or with home-manager. To Reproduce Results Running the above command: Expected behavior Syncplay installs correctly without error. Notify maintainers pyside2: syncplay: Metadata Please run and paste the result.\nThis not only affects syncplay. Many other dependencies are broken as well. Freecad for instance does not build either.\nlooks like updating pyside2 to 5.12.6 fixed this issue on my fork:\nYour fix looks good to me. Could you PR it?\ndone,"}
{"_id":"q-en-nixpkgs-0da3e4cd7e69fe1094f8854779b10f812605cbd3f0bf1b967676aa364f77f67c","text":"Describe the bug oneDNN is a dependency of pytorch. Today, oneDNN's are all linux-based but oneDNN actually appears to compile fine on ; its tests just all immediately fail because of RPATH problems. To Reproduce Expected behavior I expect oneDNN to build and check successfully. Additional context I think that this problem may be able to be fixed by adding in on Darwin, much like is currently set and exported. Note, for even more detail/context on CMake's support for controlling RPATH on macOS, please see Notify maintainers Metadata (Note, I am actually testing this using nix flakes, with this ) Maintainer information:"}
{"_id":"q-en-nixpkgs-c654bed2676bb916e6f3c1f28c8df156f91296e346bfd1f8d4096e00c861739b","text":"Describe the bug AppImages assume that the gmp library is present on the system, but appimage-run does not include gmp. To Reproduce Steps to reproduce the behavior: +x the appimage Expected behavior The app should run. Additional context I believe it is because of this commit: I fixed it on my end with: Notify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. --Metadata system: host os: multi-user?: sandbox: version: channels(root): nixpkgs: Maintainer information:\nI marked this as stale due to inactivity. &rarr;\nnot stale\nAlso affected:\nas pointed out by t-paul, appimage specifically assumes GMP is provided by the system"}
{"_id":"q-en-nixpkgs-38ccf8ca62928efa3f8c27c11d967a59c226958fe30e8e2a0b73ad18e854db2b","text":"See e.g. or\nI've been hoping for someone else to look at this one, but I expect I have to give it more visibility, hence this ticket."}
{"_id":"q-en-nixpkgs-91353d7ca5f32cdcbf5b6be557366a97bc289d30806380c9ca1f2770d929cab7","text":"The current version of qutebrowser (1.13.0, nixpkgs svn-revision: .) packaged in unstable fails to start with webkit backend. Forcing the backend in the resulting error dialog does nothing. To Reproduce Steps to reproduce the behavior: in qutebrowser Screenshots ! Notify maintainers\nThe qtwebkit backend was disabled in . Apparently at the time there was talk about dropping the backend but eventually they decided to keep it for a while. In any case here's what the current README says about it Soo, we could revert the commit but it seems it's going to be anyway. This is the reason, it looks like a variable that points to the qutebrowser executable is not actually resolved:\nThanks for the info regarding the readme and webkit. I'm pretty new to nixpks. Let me now if this would be a viable solution: and I'll make the PR. Also, where should I branch off of and what should be the target branch for my PR. It is the best I could come up with, because webkit and webengine depend on to different versions of qt5 so, we cannot easily have qutebrowser depend on both.\nYou can just revert the commit I mentioned: that should take care of also enabling qtwebkit by default, which is probably what you want. To make a PR you should target the master branch, also take a look at the .\nHmm... I don't really know, what to make of this part: in the current version it reads: and reverting just the other changes gives me build-errors.\nUhm, try something like this:\nNow I get the original import error on starting qutebrowser, when I have set to . Maybe was the right solution after all, because that at least worked."}
{"_id":"q-en-nixpkgs-b553e1ff2f04ef281f96c511256b319d10fc6365e023ad45e78cae66b587e055","text":"Dropbox does not work for me after the last update to version to 3.8.5 (commit ). The daemon starts, but no activity at all can be observed (it doesn't sync nor start the notification applet). If I start the old version that I still have in the store (3.6.9) it works. I run nixos on 32bit. I made several attempt: unlink/relink the computer from the dropbox site; restart with fresh ~/.dropbox and ~/Dropbox dirs; etc. Nothing works.\n/cc author\ndoes output anything?\nalso, which channel are you on? I just set up a nixos-unstable 32bit vm with dropbox 3.8.8 (not yet pushed) and it works fine\nIt's probably, because with 3.8.4, dropbox' qt dependency has been upgraded to 5.4, while current nixos-stable is still on 5.3 I bet you're on stable, right? Not sure if we should backport qt 5.4 to stable, or just downgrade dropbox to 3.6.9 and hope it keeps working till there is a new stable release."}
{"_id":"q-en-nixpkgs-9d3f267099c4f5ce209c9636a7a8341aae3a9866f5ae997f58d8179881366095","text":"Describe the bug fails to build when attempting to cross-compile to aarch64 on : To Reproduce Expected behavior builds without error. Metadata nixpkgs Maintainer information:\nI marked this as stale due to inactivity. &rarr;\nNot stale - this issue still reproduces on current nixos-unstable ()"}
{"_id":"q-en-nixpkgs-ede04625d2b681398e550e7ef09df4b77c9aa13d3a597a702e1a2c038b938df0","text":"Mattermost fixed a security issue in 4.5.1:\ncc\nthanks for bringing this up. I'll create the necessary PRs."}
{"_id":"q-en-nixpkgs-d089c6358e7b00e932740b59d6b7e61f73d6df07585a62739454deb0a8f32d8b","text":"Describe the bug Bazel is currently broken on the latest master, tested at . Bazel was built fine back at . I'm currently bisecting trying to find the offending commit. To Reproduce Steps to reproduce the behavior: -build . -A bazel Expected behavior Bazel builds correctly on Darwin. Additional context <details<summaryError log</summary</detailsNotify maintainers <!-- Please @ people who are in the list of the offending package or module. If in doubt, check for whoever last touched something. -- Metadata Please run and paste the result. Maintainer information:\nI suppose this might be related:\nIt's possible. I reverted , , and but I could not verify that it works because gcc failed to build. <details<summarygcc error</summary</details\nperhaps or can provide some guidance here. I'm not familiar with how it's all wired up together in nixpkgs.\nWhat’s the content of that fails?\nI'm trying to compare the environment of the working build against the one not working. As far as Nix environment variables go, there are no differences except for these that are missing: However, exporting them does not seem to make a difference.\nI've tried several different iterations, I'm unsure how to fix this. Anyone else would like to take a stab at this? Things I've tried: Adding clang to the buildInputs, nativeBuildInputs and propagatedBuildInputs. This was a long shot but gave it a try anyway. Exported the and . Added to the and of the bazelrc construction at line 421. Added various files in to NIXLDFLAGS and NIXCFLAGSCOMPILE I also compared the NIXLDFLAGS and NIXCFLAGS_COMPILE between the working version and HEAD, I did not find anything different.\nI am not quite sure what the actual error message produced by bazel. However to debug this, you can do to see what flags are passed to the actual gcc.\nThis is not set if c++ usage is not detected. See also this bit: If you want to include those manually you need to put them in instead. This can affect C programs however.\nproposed to check whether it’s a similar issue than what was solved in (He’s not available the next few days though.)\ndid fix this?\nThis should be fixed in"}
{"_id":"q-en-nixpkgs-c40a8e485681d613a3f4c0deb6a7398dc9ac790c901422340df6e5a558dabde0","text":"According to the 3.5 life-cycle () , the final release has already occurred and is now considered EOL. The jump from 3.5 to 3.6 has already happened in the majority of python packages and should not be a similar situation in which we had to extend python2's life long after its official EOL date."}
{"_id":"q-en-nixpkgs-c666e78f6e1f85d930600ccf7eafd35349ae5c47bc040881d98ef509ee1fa80a","text":"I am building nixos against the latest head and I get this warning message when running my Makefile command in my configuration It seems like it warns you about the obsolete option even after you've removed it from your configuration.\nThat's because it's used internally\nIt should be fine now, I believe. Thanks."}
{"_id":"q-en-nixpkgs-e356e7ea5f1e87c2d45de9280acbb25057441c9b3feb8dd047eefed264225483","text":"Describe the bug After fixing the the evaluation of the in I realized that it occasionally breaks with the following error: node1: output: PING 10.0.0.2 (10.0.0.2) 56(84) bytes of data. --- 10.0.0.2 ping statistics --- 5 packets transmitted, 0 received, 100% packet loss, time 4085ms error: Traceback (most recent call last): File \"/nix/store/b6m9jif1kib2dqc6fy2ym31vzs4f7hhd-nixos-test-\", line 830, in run_tests exec(tests, globals()) File \"<string\", line 1, in <moduleFile \"<string\", line 8, in <moduleFile \"/nix/store/b6m9jif1kib2dqc6fy2ym31vzs4f7hhd-nixos-test-\", line 363, in succeed raise Exception( Exception: command failed (exit code 1) cleaning up killing node1 (pid 6) killing node2 (pid 18) error: --- Error\nnix-build error: --- Error --- nix-daemon builder for '/nix/store/g14icd814yvak6syxjimgc6g8y459ki1-vm-test-run-' failed with exit code 1 Please note that the functionality isn't broken per se, on I managed to successfully build it as . Notify maintainers cc\nI'm guessing that having the pings be would resolve this, unless we want to assert that actually ensures that the peers are fully connected?\nNo it doesn't, I already tried this out, the command times out then. I guess that sometimes the isn't configured properly.\nThe issue is coming from the private key (/run/wg-priv), for some reason, it looks like networkd is sometimes trying to load it to wireguard before the systemd tmpfile gets linked."}
{"_id":"q-en-nixpkgs-1e375004a7ad1fb1437d6fdc19436e713218e4415a0cc8be24412621fa710e41","text":"Describe the bug The last several Hydra trunk-combined runs have been blocked by (among other problems) a , which was recently back(?) to release-combined in (). When I run the test locally it sometimes succeeds and sometimes hangs forever. To Reproduce Steps to reproduce the behavior: it succeeded, clear the result with (is there a better way to do that?) and try again. Notify maintainers Metadata system: host os: multi-user?: sandbox: version: channels(root): channels(anders): nixpkgs: Maintainer information:\nWe have a lot of channel blockers currently, so I recommend to also purpose revert this one. Though it sounds like it's only with issues, so perhaps comment it linking to this issue. We can always re-add it."}
{"_id":"q-en-nixpkgs-39b67c0470a1cf4f3d0dae06241c08496e35fb3a39d5284ad7d218b3020f819e","text":", CVSSv3=7.7 (nixos-unstable) CVSSv3=7.5 (nixos-unstable) CVSSv3=7.1 (nixos-unstable) CVSSv3=6.5 (nixos-unstable) CVSSv3=6.5 (nixos-unstable) CVSSv3=6.5 (nixos-unstable) Scanned versions: nixos-unstable: . Cc Cc"}
{"_id":"q-en-nixpkgs-4916fa6331c19c6c482931dcf87d76c5fee6532287051a6de5562f6b5585e43f","text":"If you want root access to an NFS share secured with Kerberos, you need to modify the default . Currently, I override the source, because NixOS has no options to modify : The text is mostly copied from the default NixOS . The only changes are This solution is brittle, and requires manual intervention if the default ever changes. I'm not sure exactly what the NixOS options should look like, but there should be mechanisms to modify in NixOS. system: host os: multi-user?: sandbox: version: channels(root): nixpkgs:\nwhat about an option like so you can easily override/append values as you see fit?\nThat looks great. I'm a big fan of RFC 42 :) Just one question though. Can you remove an option, such as ?\nGood question. can we bug you for some advice on how you can remove values from a option?\nThere's nothing builtin for that. The best that can be done is to just filter out values during string generation. Though in INI's case, is also a valid value in INI files, so that might not be a great idea. There is that adds a way to \"unset\" values, though I haven't looked into it much.\nThanks for mentioning without the ability to remove values would this not work?\nI did a double-check on , and it looks like and are actually separate, but related. defaults to the value of . Also, the default value of is already , so removing it in my configuration had no effect. On that note, we may consider removing , as that is already the default. Regardless, your RFC 42 solution should work, even without the ability to remove values.\nare you interested in opening a PR? Feel free to use the branch I posted.\nI can open a pull request. Just one question: is there a reason the option and variables are called instead of ?\nTypo\npull request is up. I fixed the typo and a syntax error."}
{"_id":"q-en-nixpkgs-2af30f448df6da1a3ed6f44e141f2488460a18b8c34446e29875aa3a0f9d3c65","text":"The package takes a long time to build the individual glyphs using . See also Other distros simply use the release tarball (e.g. AUR ) which already has the resulting 13 MB file prebuilt. Should we do this as well, given that it's a font? CC maintainer:\nI'm conflicted: on one hand NixOS is a source based distribution and building from source makes it possible to apply patches and whatnot, but on the other it's just a font without any executable stuff and 1.5 hours is a lot. I guess it depends on how valuable build time is?\nIt seems quite common for fonts in nixpkgs to use the upstream-release-provided font files; examples I picked randomly by clicking around in : Some fonts are generated by a call to , for example , but there doesn't seem to be a requirement for that. So I propose to make one of the former category.\nEDIT: confused with\nOw. Well, that's a good enough reason to switch I guess :) I'll open a pr. edit: or I could Increase the timeout :)\ncc\nI have a sentiment for source builds but no rational arguments. The above arguments for building from source are reasonable. We also have package for a bitmap version of the font (using build system from Noto). That artefact is not available upstream so we need to build it ourselves, unless we want to depend on Fedora.\nYeah, I'm similar (at least for fonts, outside fonts there's many rational arguments :rofl: ). And I think considering how it times out... It's the best choice for now as there's not a ton of benefit to argue. Though doesn't seem to be incredibly difficult to do :frowning_face.\nI think in practice this is still not going to address the root issue that the calling of inkscape is highly inefficient. Hehe, I bet nix being content addressed would make a lot of us much happier with not having to rebuild stuff pointlessly.\nI'm sorry, I confused the two packages. It's apparently the other one that's timing out on Hydra:\nEven with the 5x speedup I found in it'd still take 20 minutes to build it though. If somebody goes through the effort of fixing it upstream, we may want to consider switching back to fully-from-source. But there are 1000s of open-source tasks higher than this on my list :)\nCan relate. (says ever opensource contributor ever while submitting 10x README PRs in a day :grin: )"}
{"_id":"q-en-nixpkgs-8db2604f33b4490caf7d51724ec24195a1f5f1fa583c7bbf3aefd21d67268afe","text":"Project description DeepSpeech is an open source Speech-To-Text engine, using a model trained by machine learning techniques based on Baidu's Deep Speech research paper. Project DeepSpeech uses Google's TensorFlow to make the implementation easier. Metadata homepage URL: source URL: license: Mozilla Public License platforms: Unsure\nI already tried this with a local clone of deepspeech: But I get this error for unknown reasons:\nI guess pip is searching for a python package named sox.\nThat one doesn't seem to be in nixpkgs, is that right?\nDoesn't look like it, you might want to package that as well looks like this would need packaged as well\nThere is .\nI marked this as stale due to inactivity. &rarr;\n:(\nI binary package in , which works fine for my use case (linking to it from Rust). The binary itself works fine as well"}
{"_id":"q-en-nixpkgs-85383a592f561480f5ee9beb9b4d59574ba75d1bcee49ed42cda5cfacd7d5bdf","text":"Describe the bug To Reproduce Expected behavior It should give me a shell Additional context I can fix it with which is based off 20.09-beta but I don't seem to be able to use HEAD (it breaks for another reason). Notify maintainers Metadata Maintainer information:"}