text
stringlengths 1
5.2k
| labels
listlengths 1
1
|
|---|---|
sudo podman run --rm --privileged quay.io/libpod/alpine_labels:latest cat -v /proc/self/attr/current WARNING: The same type, major and minor should not be used for multiple devices. Something is wacky with --privileged. Looks like a bug.
|
[
1
] |
Almighty cleanbot smashes even GODMODE mice and cockroaches.
|
[
1
] |
visualization py is an example i think it is better to add a colon i am not sure it is a correct sentence
|
[
0
] |
you are mistaken you mixed generate physnet vlan ranges output with generate physnet mtus output
|
[
0
] |
missing partial bug
|
[
0
] |
s worker node jenkins slave instance
|
[
0
] |
Guys, this is really nonsense. I don't know why you did this.
|
[
1
] |
this should not be needed get project raises projectnotfound for an invalid project id
|
[
0
] |
is trash another top level drive directory like team drives and root or does each of those have it is own trash directory like team drives trash and root trash if it is the former then this lgtm if it is the latter we should add an explicit check for this
|
[
0
] |
i am pretty sure you can have exception perform a traceback logging if so then this is going be really ugly in the log file
|
[
0
] |
destroy drm plane cleanup
|
[
0
] |
Wait changelings can eat people whole? Why remove that? Horror mode is shit hitting the fan
|
[
1
] |
yes there is a bug here thank you will fix in the first place i do not understand why it is pop here instead of get stupid cosmic rays
|
[
0
] |
this should be delete oper delete
|
[
0
] |
remove juno it is eoled already
|
[
0
] |
I'm for the proposal. Many posts have a short relevance in terms of content and are then only data rubbish. I suggest to make the deletion rules individually definable for each user in the settings. This will reduce data garbage, increase privacy and, if necessary, mark important posts as permanent!
|
[
1
] |
they re now the same and i have changed the name to reflect that
|
[
0
] |
English on this first sentence is a bit weird. Maybe: The only limit on the number of loaded SoundEffects allowed is the available memory
|
[
1
] |
checkimage should be below buildstatuspanel loadsupport buildimageviewer buildstatuspanel checkimage
|
[
0
] |
btw this likely merits an aloge and possibly an immediate abort of the keystore daemon as it indicates a bug in the code
|
[
0
] |
there should be a default domain id
|
[
0
] |
they are different deploy devstack sh is on devstack slaves this one is for devstack parent host
|
[
0
] |
the copy destination should be savanna conf dir savanna conf file
|
[
0
] |
this was not even a discriminated union ugh clearly c programmers at work
|
[
0
] |
i'd just like to say I fucking love this btw best virus ever
|
[
1
] |
Wacky. Could you get a GE capture?
|
[
1
] |
wow elif should be removed at all with todo comment as it is now it skips all extra fk that was found
|
[
0
] |
That's what I argued at the time because it totally looks like we screwed up somehow. The fact that we're surfacing the `undefined` type isn't a problem so much as it is that its representation kinda, well, sucks. :smile:
|
[
1
] |
i think we change to oslo utils everywhere
|
[
0
] |
usually the pollsters in ceilometer interface directly with the python client is that going to be case as well does ceph has a python client
|
[
0
] |
wtf is this magic number in the first place these should probably at least be modifiers instead of returning a whole value (like this returns .3 which is added to 1 later on)
|
[
1
] |
first of all this breaks indentation second this fuction is a little useless see below
|
[
0
] |
iqn self test vol
|
[
0
] |
Killer tomatoes should be 5-10 damage at most probably - they aren't exactly very scary critters
|
[
1
] |
should we still log a meaningful message if the cluster was not found instead of having an attributeerror
|
[
0
] |
What a disgusting comma splice.
|
[
1
] |
instead of the mock why not just call manager neutronmanager set controller for resource and pass it your fake controller
|
[
0
] |
i would still prefer u use l instead of a raw greenthread
|
[
0
] |
Weird indenting. Are there tabs used?
|
[
1
] |
Damn. I keep doing that.
|
[
1
] |
check that num volumes is an int that is greater than
|
[
0
] |
blocked without sniffing maybe maybe reorder
|
[
0
] |
os brick encryptors use castellan and this bp will call the library os brick encryptors
|
[
0
] |
@OKNoah You're kinda coming off as a jerk. Just like you, @martinwoodward is a person...a human being. Labeling someone as a loser isn't going to bring yourself any respect or change anyone's opinion. It only makes other humans tune you out.
|
[
1
] |
> I was drunk as hell when I coded this Haha yeah I get it
|
[
1
] |
I had that recommended to me and I couldn't for the life of me get that working without fucking the roburger up.
|
[
1
] |
extra fk
|
[
0
] |
The function should probably be renamed back to `is_disgusting_for()` and accept an argument (just like it did earlier).
|
[
1
] |
This is awful and honestly I'd be less direct if you didn't spend an entire paragraph flaunting about how good you are and very little to explain the reason behind these changes There are no unspoken rules of mapping, if you can't justify a change it's probably a bad idea. EOBGames pretty much nailed it, emphasis mine on the fact that the doors are there to stop randoms from barging in during treatment which was one of the main issues with old medbay
|
[
1
] |
is the static cast const actually necessary and is duplication of a one liner really that bad const cast is really ugly no matter where you put it i would rather see the logic of getting the vector element duplicated than the const cast
|
[
0
] |
Yeah, this proc is awful and full of spaghetti and everything sucks, but oranges' solution sucks slightly less than yours.
|
[
1
] |
this is still lvm iscsi also below please change
|
[
0
] |
hmm if deprecated name deprecated group and deprecated opts are all none then i would expect deprecated opts to be empty
|
[
0
] |
yup
|
[
0
] |
God, just define `ITransformComponent` as a dependency when you make this use resolves and don't do this shit
|
[
1
] |
yes totally done the intention was to let this zone die as soon as possible but the whole liftoffcompiler will die after compilation anyway
|
[
0
] |
qos id qos db item id
|
[
0
] |
@chrissPony Yes, sadly 2.x has some wacky and inconsistent inflection of path parts which cannot be changed due to backwards compatibility reasons. Just don't want to add more to it :)
|
[
1
] |
A gorrilion commits and a awful change, sign me up for some drama.
|
[
1
] |
system shell execute uses processutils execute using subprocess popen i want to complete this asynchronously
|
[
0
] |
fc related code should be removed
|
[
0
] |
> ridiculous is quite good and less insulting Thank you. I replaced "stupid" to "ridiculous".
|
[
1
] |
Give credits fag
|
[
1
] |
why not mock object same for another mocks below
|
[
0
] |
operatortest factorytest
|
[
0
] |
there seem to be many uses of jsonobject within chrome i am not sure if any are parsing web content though there is good support on the native side for parsing json in a utility process so it is not a big hurdle if you prefer that the structure is currently just or so ntp entries which each contain an id string icon url if we go utility process from then this is kinda moot
|
[
0
] |
remove pyc
|
[
0
] |
i took a stab at rewriting this description and make this more clear
|
[
0
] |
Saw this happen in game and it was kinda disgusting, really exploity. Seems this has already caught on as the hard counter meta. I really support this PR now
|
[
1
] |
this leads to jobs p kill kill usage kill s sigspec n signum sigspec pid jobspec or kill l sigspec if there is no background jobs
|
[
0
] |
shit youre right, committing fix now
|
[
1
] |
ahh well i will see your target simulator and raise you a board uses generic audio which does nothing as where it was used is commented out it seems seeing as random makefile cleanup seems to be my current calling i shall no doubt take a look we re also only a couple of commits away from killing libunz too
|
[
0
] |
the should probably be lined up
|
[
0
] |
does this produce any useful output you are crashing a working kernel thread
|
[
0
] |
if it is for sure an error related to trying to delete an nonexistent route the log could say so
|
[
0
] |
see dictionaries lists this one is preffered better to change all other
|
[
0
] |
i think you should check you did eliminate the instructions you think should be eliminated
|
[
0
] |
I have now added it to my own bintray, but this is a bit ridiculous. I have no knowledge on the package itself. It's still awaiting inclusion in jcenter but maybe it would be easier to include it in openhab/mvn.
|
[
1
] |
probably want to kill this comment
|
[
0
] |
if streaming will be handled in nova could you wrap the operation using zlib perhaps to compress the data sorry if this is a stupid question
|
[
0
] |
to be pedantic i comes after h
|
[
0
] |
these eqlx are vendor specific promoting this to the base does not look good is there another way
|
[
0
] |
Ah, I get it now! Thank you so much; you're as brilliant as ever, @diddledan. Before I resign myself to silently spectating, I feel I ought to point out how ridiculous & absurd this is? It feels like adding more layers of duct tape isn't the approach Snapcraft should be taking? Perhaps I'm just too used to the more consumer-friendly world of Windows, but forcing developers to jump through all sorts of hurdles just to publish their rather simple software is highly concerning, and presents a major issue in promoting FOSS development. For the few like me who're willing to deal with such things, there are many who will be turned off from the whole ecosystem entirely. Anyway, it's really not my place, I suppose. Again, thank you for the continued work. You lot are far more talented than I.
|
[
1
] |
i think john meant subprocess pipe then you can use map output stderr
|
[
0
] |
i think that is fine bionic internal code should probably be using lock anyway
|
[
0
] |
this looks wrong you have self mocks other arguments you are missing a mock
|
[
0
] |
urlencode again would be nice instead of doing the string munging yourself
|
[
0
] |
self f should be self test file
|
[
0
] |
actually the client is intended to be as dumb as possible we might want for example to test passing extra data
|
[
0
] |
silly nit but put this inside the if
|
[
0
] |
yeah i am an idiot i was comparing to the my workspace eventfilter which did not have your changes
|
[
0
] |
unless gerrit s lying you left a double space here
|
[
0
] |
i am not spending time on that here because the future will do this for us it would require careful auditing of every call point to make sure we do not break things i want to kill optparse first before moving on to autofreeze
|
[
0
] |
ha true i am an idiot
|
[
0
] |
you can drop these die
|
[
0
] |
input file will be undefined i believe you need pep input file
|
[
0
] |
the const cast is evil can t you just look up the axnode from tree by id and then modify that
|
[
0
] |
what the fuck is that
|
[
1
] |
should not the the style here be warning or are you letting a formatter take over
|
[
0
] |
you need learn puppet about munge funcion for example
|
[
0
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.