text
stringlengths 1
5.2k
| labels
listlengths 1
1
|
|---|---|
looks like you have a bad rebase because live migration copy disk paths only takes a single argument now so we have this in the multinode jobs for live migration testing
|
[
0
] |
any takers on fixing i can take a stab at it tomorrow
|
[
0
] |
nobody uses createperusertopicinvalidationclient who is going to own invalidationclient does it make sense to return std unique ptr invalidationclient
|
[
0
] |
No, don't fuck with body temp, find a different way to suppress that slowdown
|
[
1
] |
according to hacking there should be a blank line here
|
[
0
] |
wow, cyclops are people too, rude
|
[
1
] |
Notable stupid shit found in blob_act() If there was no fireaxe in the fireaxe cabinet, it would not be deleted by the blob.
|
[
1
] |
pass is useless
|
[
0
] |
there are time like this when i found myself to be really ignorant can you explain me why one should use setdefault here
|
[
0
] |
midonet client midonet neutron client api midonetapiclient midonet use zoom false
|
[
0
] |
please remove halign start to align the control left and right
|
[
0
] |
should there be either of the arguments u force reinstall
|
[
0
] |
these configs should appear in etc ironic ironic conf sample to update that file do something like tools config generate sample sh b p ironic o etc ironic
|
[
0
] |
man, this is ridiculous; I wonder why it happens 
|
[
1
] |
Why the fuck is it even using :? It's already typecast.
|
[
1
] |
- We had some wacky conflicts with master. Rebased and resolved conflicts, let's see if CI pass. - Remaining work is to update buckets based on feedback \u2013 will do this next week.
|
[
1
] |
why are you mocking this i think you should be mocking the instance update method or even just the save on the instance object maybe you are still passing a sqlalchemy model to those methods when you should now be passing a novaobject and mocking that save
|
[
0
] |
this backslash is unnecessary and should be removed
|
[
0
] |
the options you moved from service instance py to here i e service network cidr are no longer available in the manila conf sample file you need to import manila share drivers service instance helpers neutron helper in manila opts py and add manila share drivers service instance helpers neutron helper helper opts to global opts list you can try the result with tox egenconfig
|
[
0
] |
src map nullptr
|
[
0
] |
the issue is that this is keepalived pm and not vrrp pm which is the problematic one it s very weird by keepalive to have such design with separate pid files we end up needing to watch i hope it will not be common if it becomes common i would also expect a more general solution
|
[
0
] |
> But it is indeed wacky hidden automagic which should be changed in 3.6. How do we go about removing the current behavior without breaking existing applications? Removing the id addition seems like the kind off thing that would break userland code.
|
[
1
] |
Do we still want to use those new macros for single bit checks (flags)?
|
[
1
] |
> Also 1 second is a VERY long time to upload telemetry in an average case (I would have expected 1/10th that). Please bear in mind users (such as myself) located on the other side of the world with crappy internet, high latency and heavily constrained connections.
|
[
1
] |
why use kwargs here at all this is the final func so people should not be passing in junk at this point sync instance should be a proper arg
|
[
0
] |
You should read your code again and be ashamed. _also that piss poor copypaste smh_
|
[
1
] |
Why the fuck did you put a spawnpoint in space you stupid asshole
|
[
1
] |
should not this go before brctl addi fuck
|
[
0
] |
are unnecessary and should be dropped
|
[
0
] |
pre version does not tell us much better would be test upload to image pre
|
[
0
] |
That works, I figured that people would not understand what the concept of categories is, but I suppose they shouldnt be fucking with the code if they dont want to go find out themselves.
|
[
1
] |
can you remind me what we do with the zygote non moving space coming right after the boot image oat if we do not include the oat if there is no image after then could we miss including the zygote non moving space in the immune region do we have a test for a dummy zygote non moving space
|
[
0
] |
no one in the right mind will set retries to or come on that said this is correct to me db max retries i should not run and it is consistent db max retries i retry only one and i shoot myself in the foot i really do not see the problem here but it could be just me
|
[
0
] |
god catch this was a rebase error from my side thanx will upload another patchset
|
[
0
] |
needs to be a nul terminated string
|
[
0
] |
we re talking about a u so i would write this simply as
|
[
0
] |
could you move the mock execute code into setup function for example p mock patch object self driver execute self execute p start self addcleanup p stop self execute assert called
|
[
0
] |
please get rid of the trailing space
|
[
0
] |
the surl isempty part is redundant now and should be removed
|
[
0
] |
these args are not needed and should be removed
|
[
0
] |
this seems redumnant tm i do not believe that you need to set culsters manager mock mock because you want to set clusters manager delete
|
[
0
] |
Are you serious? Goddamnit I spent the most time on this one.
|
[
1
] |
whatsa hardwark
|
[
0
] |
yuzhu so i took a look at where i would think the documentation would be on guarantees lack of guarantees of state after a watcher was signaled with a readable signal watching handles and did not find anything that clarified this is there somewhere else i should be looking given that we have some code that appears to assume that a watcher triggering with a readable signal can always be read and some that doesn t it would seem making sure the documentation is clear about what assumptions can be made willing to give me a pointer to the documentation i should be considering updating
|
[
0
] |
ideally the undef would not be quoted undef is a special puppet value like nil in python
|
[
0
] |
to sum things up are we sure that e will be an instance of httpexception yes i am sure that e is an instance of httpexception or it is subclasses
|
[
0
] |
Oh crap I erased the head by accident :/ I think it's not a big deal
|
[
1
] |
can you chk if we have any negative tests with ttl iirc we do not have any but in case we do can you remove them
|
[
0
] |
should we use a constant here instead of a naked
|
[
0
] |
i will need to remove from agent ucs driver which i am doing in the next patch
|
[
0
] |
that s like creating an issue on github so i can pr very silly i will do it though
|
[
0
] |
though this does not seem to be the issue with my local... theres something wacky about the layout on mine
|
[
1
] |
wrapping swdev id should be detected and stopped here
|
[
0
] |
please remove these extra white spaces
|
[
0
] |
ofport self del fdb flow br port lvm cleanup tunnel port br ofport lvm network type add above two lines which could finish the garbage collection before add eliminating a separate fdb delete rpc
|
[
0
] |
Damn it I always forget about this, will fix. But God I hope not.
|
[
1
] |
Gosh darn, it looks like the bit definitions aren't even compatible, for comparison with glibc https://code.woboq.org/userspace/glibc/bits/dlfcn.h.html
|
[
1
] |
silly nit i would prefer snapshot instead of snap
|
[
0
] |
don t you think it would be better to pass this option through env variable or through run tests sh i mean i think it should not be hardcoded
|
[
0
] |
i think it does not make big point
|
[
0
] |
why not like create backup function use greenpool spawn n to execute delete backup function
|
[
0
] |
yes? it's still used internally for processing, read the fucking description you dickwad
|
[
1
] |
Curse you, autocorrect!
|
[
1
] |
the newest commit should be 5 minutes, unless i fucked up my math
|
[
1
] |
Hot damn, this entire file may be the worst code I've ever read.
|
[
1
] |
is not it simpler to use self vpnservice with do delete false and catch webob exc httpclienterror
|
[
0
] |
why not just call this absoluteresource to avoid the need for the odd dummy class below
|
[
0
] |
useless setup
|
[
0
] |
junk comment
|
[
0
] |
:-1: if you die you shouldn't lose the experience, fuck off with this shit its about learnning to play not whether you lost those 2 hours of experience in a freak accident
|
[
1
] |
use named string interpolation instead of positional
|
[
0
] |
hmm fixed ips fixed ips
|
[
0
] |
kirill you are wrong this command kills only parent process instead of killing all processes like pkill does
|
[
0
] |
so this is a dumb question i am not sure who what calls spawn but it is not this driver and this is in nova should we be raising ironic exceptions here or converting them to nova exceptions
|
[
0
] |
i am all for gatting up the fools who step to us but perhaps you meant investigate
|
[
0
] |
stealth changes or git fuckery?
|
[
1
] |
what is about real path readlink f
|
[
0
] |
> Just ran the script and there was no crash. However, the fits are rubbish. Is that expected for this example? > > Also in the parameter table, which values should be the same? Yes, I would not expect good fits for this example. It is just a very simple example from the unit tests with very few iterations so it does not take too long to run it. Some parameter like FWHM will not change much (if at all) after such a short time.
|
[
1
] |
i copied this shamelessly
|
[
0
] |
to be removed here and below
|
[
0
] |
don t we need to run a different command if that haproxy is alreday running e g some kill cat pid
|
[
0
] |
let s leave this as valueerror
|
[
0
] |
Ashwalkers are fucked if they spaen on a island and cant cross lava. Hell golems too really
|
[
1
] |
lgtm small thing is the update and delete needed to finish the crud
|
[
0
] |
useless trailing see the implementation of the rand name function
|
[
0
] |
just want to point out there is a separate mock object for mocking open mock open
|
[
0
] |
dolph actually the docstring here means resource to validate is not existing in the kwargs
|
[
0
] |
Ridiculous that people are asking for features and bug fixes on some other unrelated repo just because Microsoft doesn't do it themselves and because the Feedback Hub is awful. But yay, this was added \U0001f389
|
[
1
] |
i think you should also add a bullet that says improve test coverage
|
[
0
] |
this is the existing sigquit dump thing which dumps the existing gc performance data
|
[
0
] |
yes addr create input chipfamily family cz
|
[
0
] |
because of absents of ability to delete security group attached to instance if no server check we could catch punch in a face with exception while deleting security group
|
[
0
] |
especially considering these stack, do not make eating without a table, chair, and forks make you instantly fucking depressed/insane
|
[
1
] |
yikes silly me you re right i missed that change somehow could we update the docstrings of such methods to explain that if a port is more than once in the list that means we expect the event to happen twice or three or n times i guess that is the current behaviour right thanks for the patience rossella
|
[
0
] |
Goddamned Sony and your regional variants.
|
[
1
] |
leave out until reach weapons are unfucked then make it a super cool reach weapon
|
[
1
] |
this conditional section is pretty ugly
|
[
1
] |
done ah thieu you will not let me be lazy will you
|
[
0
] |
Fucking, Masks are defined as actual masks, most other slots are /item. looks like it isn't possible (unless you repath mask code to not assume type, like all other slots)
|
[
1
] |
Woah what happened to the Git diffs! 1000+ changes. Something went wacky here.
|
[
1
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.