showvikdbz/code-reviews · Datasets at Hugging Face

text stringlengths 1 5.2k	labels listlengths 1 1
this is obviously a dirty formatting and this is probably a pourly choosen tag name	[ 0 ]
> Gives you a nasty case off Tourette's > off you fucking kidding me?	[ 1 ]
you used plain type again here fwiw above you fixed one by not having a local variable and the other by renaming the local variable either works but it is a tiny bit nicer to be consistent	[ 0 ]
somehow i managed to miss the rest of this thread i wonder if i fat fingered the mute button in my reader anyway i tend to agree with you maybe this check is not worth adding	[ 0 ]
in model you have not added this column this caused failure of neutron dsvm functional tests also do not we need an fk on this column	[ 0 ]
use meaningful names table argument index priority argument index first call args self l app add flow normal call args self assertequal first call args table argument index self assertequal first call args priority argument index	[ 0 ]
How is `boring` any more or less applicable to this list than `evil`, `high`, `goofy`, `tender`, etc? Everybody's boring sometimes (except Woz, but that's the joke).	[ 1 ]
Ah, I'm glad those aren't the final numbers. I know the old numbers could get ridiculous, but it was fun and made grinding skill levels more rewarding.	[ 1 ]
do we need to code around this can we just kill the networkinterfacefactory and generate it like the rest attach interfaces to driver could then use interface loaders network	[ 0 ]
sorry for the ignorance is it at all possible to run mcollective at all with a non root account	[ 0 ]
clarification question dumb question will the policy criteria always allow for creating a valid symmetrical patch successfully	[ 0 ]
nit delete already tests for nullptr so this is equivalent to delete sinstance sinstance nullptr	[ 0 ]
first i adopted your suggestion honestly i was hesitating between what you suggested and the current version when i implemented this days ago if process isvalid it could be a direct child of iter second but has been killed resulting from it is parent process being killed correct it could have been if base process open fails on the other hand if base process open succeeds but finds a proc with a different parent due to pid reuse we know that we should ignore this process however it is child processes may not be killed yet so we need to add it is pid to the flat map and continue hunting for the descendant processes ah i see that is a good point but if the pid was reused we do not want to kill the new proc s kids to be totally paranoid openprocessifchildof would need to report to callers whether the pid was definitively reused or not i think what you have now is good enough for test code maybe drop a comment explaining the possibilities since it is a bit subtle either version works fine for this specific case as killing the direct child process automatically kills all the descendant processes gpu utility etc per my local tests yes chrome puts it is sub procs in a job object so that they all should die with the parent at least i think that is how it works you can probably test this trivially by killing a browser proc with plenty of subprocs i think they evaporate straight away	[ 0 ]
literal af literal	[ 0 ]
These still look failed to me, plus the code coverage numbers are kinda wacky.	[ 1 ]
hmm you are setting prebuilts to false first then to true later you can probably just take out both lines	[ 0 ]
is the code catching typeerror just in case pid none it would read better if you just said this up front then some of this code could be cleaned up if pid and self root helper child pid none while true try child pid utils find child pids pid except indexerror last process in the tree return it return child pid	[ 0 ]
Whoops, forgot to remove that one when cleaning up shitcode	[ 1 ]
Hahah, I hate when the poop is repeating! ```suggestion /// If the pool is repeating ```	[ 1 ]
ha i was looking for the dcheck to add here but i believe it is sitting right in front of my eyes i ll have to add a version that is assert blocking allowed so my feeble brain can see it	[ 0 ]
I guess I copy pasted in some darned fancy dash or something. Amended.	[ 1 ]
I\u2019m a bit more reluctant to remove `driver` since it might be a useful search term. Drivers have such ridiculous names that it may be a good way to filter what you\u2019re looking for from the cruft.	[ 1 ]
Yeah but that would be OOP who the fuck wants that \U0001f602	[ 1 ]
Now you have the info. In L4 we got a precise Exception, and now we get this rubbish...	[ 1 ]
Yes, ghosts cant do shit by default, no need to add ghost checks	[ 1 ]
remove mock clients then	[ 0 ]
pedantic consider the following port fips floating ips get port id for fip in port ips	[ 0 ]
Damn it, I didn't proof read that part	[ 1 ]
Windows folks are so insensitive!	[ 1 ]
no i think this is preferable to have an informational output because a failure	[ 0 ]
We cannot `STRINGIFY` code in C++. Only preprocessor defines.	[ 1 ]
this whole chunk is a lot of new code i do not see where it is tested at all in fact the only test adds in this patch are a few mocks to make some current tests still pass	[ 0 ]
i m sorry if that is a dumb comment i wonder whether those really belong here should they be maintained in service repos instead	[ 0 ]
can you document what the hell this proc does, because I do not understand by looking at it	[ 1 ]
please check the error message returned	[ 0 ]
poolornew is shit, use new	[ 1 ]
this is such a nit i feel like an idiot even mentioning it if not x with an else always bugs me i always have to double take by the time you get to else because it is a double negative this pattern is below and not something you need to change just a rant	[ 0 ]
when i was using ndk root platforms some needed files were not found apparently we have not really root caused this failure	[ 0 ]
dumb question alert why this is not utils get nameservers none none	[ 0 ]
oh damn, we getting serious over here	[ 1 ]
Seems packit is completely b0rked right now.. But also, [this](https://cockpit-logs.us-east-1.linodeobjects.com/pull-17705-20220916-151134-e258e1fd-fedora-36-pybridge/log.html) is completely ridiculous right now.	[ 1 ]
the colon makes this invalid yaml you can either quote the whole thing or do something like other the blah blah manage max workers	[ 0 ]
i am pretty sure you should use instead of in function names i think dashes are a bashism	[ 0 ]
i have pulled traversal into ngpaintfragmenttraversal	[ 0 ]
i asked for this because i like less coupling with the conf object	[ 0 ]
another one although i think killing the extra space is okay	[ 0 ]
tab only lines are shit anyway	[ 1 ]
because args is a tuple	[ 0 ]
done greg and rob had a lot of discussion on whether it should be null terminator or string terminator let us use null string terminator something suggested by rob here to make everyone happy	[ 0 ]
I hate being unable to suggest changes on lines that have "been deleted", screw you github. ```suggestion /datum/action/item_action/mod/Grant(mob/action_receiver) ```	[ 1 ]
@rwjblue it seems wrong that there is a placeholder script tag at all. I can look into that. This is a little wacky, but we can probably do better.	[ 1 ]
another bogon dead	[ 0 ]
That's funny, so the "ERROR: " were the actual inconsistency. My next push will have a lot of formatting changes, even if it's not exactly exhaustive because... oh my goodness, this file is... ridiculous	[ 1 ]
now its a font shitting up your browser	[ 1 ]
WTF, sorry i new on github.... some misclick	[ 1 ]
you can use assertin and assertnotin	[ 0 ]
iirc dumb mapping of non dumb buffers is not supported it happens to work on some drivers but we should probably not rely on it too much	[ 0 ]
this is such an irrefutably stupid, awful, and just downright bad idea but it's so fucking funny in theory that it just might work...	[ 1 ]
Yea, what i originally put in but the sprite was fucky when testing, prob how the old file was formatted so I'll change it	[ 1 ]
Why are there so many issues with angular 8? Its getting ridiculous	[ 1 ]
JFC @cpojer fix jest. This is ridiculous. @spicyj - did you by chance test with iojs+jest@0.5 first?	[ 1 ]
Damn, we need to support switching global css asap :)	[ 1 ]
This sound still makes me physically cringe whenever I hear it	[ 1 ]
seems I somehow screwed it up :see_no_evil:	[ 1 ]
Hey, Hugh, I'm not wild about the line length changes that autopep8 is making. The line breaks are pretty clumsy, and I think they make the code harder to read in many cases. Can you turn particular change off?	[ 1 ]
the ridiculous highlighted error line is here, since it seems highlighting changes after job completition (I assume due to pod logs vs the uploaded logs?) https://gist.github.com/BenTheElder/691ec52da5641df109374b5ba133febd	[ 1 ]
This is really bizarre. What happens when you load https://keybase.io in a browser?	[ 1 ]
Damn fancy acronyms, who did that?	[ 1 ]
stack sh is run under a mere user so i think yes	[ 0 ]
i meant fk with on delete cascade	[ 0 ]
Goofball doesn't want the normies to have his cool tools.	[ 1 ]
i think the part about the user is responsible can be removed block on l will be executed anyway as i read it	[ 0 ]
need tenants more api keystone tenant list request domain none paginate false	[ 0 ]
i think we can kill this using jsonpath rw ext	[ 0 ]
this seems to me to be another id id problem did we ever decide the proper usage of id in the context of cloudcafe	[ 0 ]
uh i am assuming these defaults are really just dummy values you should probably remove them	[ 0 ]
this init is useless	[ 0 ]
this exec should be removed we do not need to minify	[ 0 ]
> making blob weaker vs fucking every obstacle FUCKING STOP	[ 1 ]
gay code	[ 1 ]
okay took another pass at this and made the entirety of the promise tracking significantly simpler previous code was copied from webvr path which i did when i was young and stupid a whole years ago	[ 0 ]
does people really find out this hair swag? Because it's fucking disgusting.	[ 1 ]
so this whole check is bunk i only have mocks here because i override the side effect on with one on so this actually fails properly when i check build req destroy called	[ 0 ]
nit i think this should be just another bullet point to join the above list	[ 0 ]
Sorry if this is ugly; kind of hard to initialize a local due to scoping rules.	[ 1 ]
a language that does not have audio plugin of its own seems ridiculous. Still waiting for the flutter team to come with a solution.	[ 1 ]
i think i am missing something here but i do not know this code at all wall so sorry for the dumb questsion if we fallback to the integration bridge i would have expected to see this tried first then if not found attempt to look up the integration bridge	[ 0 ]
you are addict to this construction though it is not a must here i would say it makes reading harder	[ 0 ]
w is a warning that a file should end in a blank newline h is the hacking rule about always having a single char summary of all docstrings something i personally think is silly and since nova also ignores it l i figured we could as well	[ 0 ]
include another test in here (good practice to make sure loops can't go forever even if some dumb shit happens)	[ 1 ]
Dude this is huge. Just edit it and keep the relevant parts. Add only the first few days after genesis. Delete all irrelevant data. Delete all irrelevant html stufff. Do not commit useless stuff in the repo	[ 1 ]
error message should state which module to import the function from, it's ridiculous not to see that on the development console.	[ 1 ]
isn t this what management false is supposed to do i think you can revert this make sure you haven t broken the admin public mechanism	[ 0 ]
yeah i considered that but i also wanted to instantiate this class as a fallback when the configure inspector type cannot be loaded for any reason	[ 0 ]
use die instead of this and the printk	[ 0 ]
An atmos pipe has a height of 1m and a radius of approximately 15cm In freedom units that's a height of about 1 yard and a diameter of about 1 foot You don't want that shit inside of you	[ 1 ]
Er, this could fuck up GCing on ID cards.	[ 1 ]
this needs a docimpact tag please thanks	[ 0 ]
the backslashes were perfectly indented before	[ 0 ]
edit also i am an idiot and it should of course be bytes per pixel in the framebuffer for rgb so same question but make the math and align up	[ 0 ]

this is obviously a dirty formatting and this is probably a pourly choosen tag name

[ 0 ]

> Gives you a nasty case off Tourette's > off you fucking kidding me?

[ 1 ]

you used plain type again here fwiw above you fixed one by not having a local variable and the other by renaming the local variable either works but it is a tiny bit nicer to be consistent

[ 0 ]

somehow i managed to miss the rest of this thread i wonder if i fat fingered the mute button in my reader anyway i tend to agree with you maybe this check is not worth adding

[ 0 ]

in model you have not added this column this caused failure of neutron dsvm functional tests also do not we need an fk on this column

[ 0 ]

use meaningful names table argument index priority argument index first call args self l app add flow normal call args self assertequal first call args table argument index self assertequal first call args priority argument index

[ 0 ]

How is `boring` any more or less applicable to this list than `evil`, `high`, `goofy`, `tender`, etc? Everybody's boring sometimes (except Woz, but that's the joke).

[ 1 ]

Ah, I'm glad those aren't the final numbers. I know the old numbers could get ridiculous, but it was fun and made grinding skill levels more rewarding.

[ 1 ]

do we need to code around this can we just kill the networkinterfacefactory and generate it like the rest attach interfaces to driver could then use interface loaders network

[ 0 ]

sorry for the ignorance is it at all possible to run mcollective at all with a non root account

[ 0 ]

clarification question dumb question will the policy criteria always allow for creating a valid symmetrical patch successfully

[ 0 ]

nit delete already tests for nullptr so this is equivalent to delete sinstance sinstance nullptr

[ 0 ]

first i adopted your suggestion honestly i was hesitating between what you suggested and the current version when i implemented this days ago if process isvalid it could be a direct child of iter second but has been killed resulting from it is parent process being killed correct it could have been if base process open fails on the other hand if base process open succeeds but finds a proc with a different parent due to pid reuse we know that we should ignore this process however it is child processes may not be killed yet so we need to add it is pid to the flat map and continue hunting for the descendant processes ah i see that is a good point but if the pid was reused we do not want to kill the new proc s kids to be totally paranoid openprocessifchildof would need to report to callers whether the pid was definitively reused or not i think what you have now is good enough for test code maybe drop a comment explaining the possibilities since it is a bit subtle either version works fine for this specific case as killing the direct child process automatically kills all the descendant processes gpu utility etc per my local tests yes chrome puts it is sub procs in a job object so that they all should die with the parent at least i think that is how it works you can probably test this trivially by killing a browser proc with plenty of subprocs i think they evaporate straight away

[ 0 ]

literal af literal

[ 0 ]

These still look failed to me, plus the code coverage numbers are kinda wacky.

[ 1 ]

hmm you are setting prebuilts to false first then to true later you can probably just take out both lines

[ 0 ]

is the code catching typeerror just in case pid none it would read better if you just said this up front then some of this code could be cleaned up if pid and self root helper child pid none while true try child pid utils find child pids pid except indexerror last process in the tree return it return child pid

[ 0 ]

Whoops, forgot to remove that one when cleaning up shitcode

[ 1 ]

Hahah, I hate when the poop is repeating! ```suggestion /// If the pool is repeating ```

[ 1 ]

ha i was looking for the dcheck to add here but i believe it is sitting right in front of my eyes i ll have to add a version that is assert blocking allowed so my feeble brain can see it

[ 0 ]

I guess I copy pasted in some darned fancy dash or something. Amended.

[ 1 ]

I\u2019m a bit more reluctant to remove `driver` since it might be a useful search term. Drivers have such ridiculous names that it may be a good way to filter what you\u2019re looking for from the cruft.

[ 1 ]

Yeah but that would be OOP who the fuck wants that \U0001f602

[ 1 ]

Now you have the info. In L4 we got a precise Exception, and now we get this rubbish...

[ 1 ]

Yes, ghosts cant do shit by default, no need to add ghost checks

[ 1 ]

remove mock clients then

[ 0 ]

pedantic consider the following port fips floating ips get port id for fip in port ips

[ 0 ]

Damn it, I didn't proof read that part

[ 1 ]

Windows folks are so insensitive!

[ 1 ]

no i think this is preferable to have an informational output because a failure

[ 0 ]

We cannot `STRINGIFY` code in C++. Only preprocessor defines.

[ 1 ]

this whole chunk is a lot of new code i do not see where it is tested at all in fact the only test adds in this patch are a few mocks to make some current tests still pass

[ 0 ]

i m sorry if that is a dumb comment i wonder whether those really belong here should they be maintained in service repos instead

[ 0 ]

can you document what the hell this proc does, because I do not understand by looking at it

[ 1 ]

please check the error message returned

[ 0 ]

poolornew is shit, use new

[ 1 ]

this is such a nit i feel like an idiot even mentioning it if not x with an else always bugs me i always have to double take by the time you get to else because it is a double negative this pattern is below and not something you need to change just a rant

[ 0 ]

when i was using ndk root platforms some needed files were not found apparently we have not really root caused this failure

[ 0 ]

dumb question alert why this is not utils get nameservers none none

[ 0 ]

oh damn, we getting serious over here

[ 1 ]

Seems packit is completely b0rked right now.. But also, [this](https://cockpit-logs.us-east-1.linodeobjects.com/pull-17705-20220916-151134-e258e1fd-fedora-36-pybridge/log.html) is completely ridiculous right now.

[ 1 ]

the colon makes this invalid yaml you can either quote the whole thing or do something like other the blah blah manage max workers

[ 0 ]

i am pretty sure you should use instead of in function names i think dashes are a bashism

[ 0 ]

i have pulled traversal into ngpaintfragmenttraversal

[ 0 ]

i asked for this because i like less coupling with the conf object

[ 0 ]

another one although i think killing the extra space is okay

[ 0 ]

tab only lines are shit anyway

[ 1 ]

because args is a tuple

[ 0 ]

done greg and rob had a lot of discussion on whether it should be null terminator or string terminator let us use null string terminator something suggested by rob here to make everyone happy

[ 0 ]

I hate being unable to suggest changes on lines that have "been deleted", screw you github. ```suggestion /datum/action/item_action/mod/Grant(mob/action_receiver) ```

[ 1 ]

@rwjblue it seems wrong that there is a placeholder script tag at all. I can look into that. This is a little wacky, but we can probably do better.

[ 1 ]

another bogon dead

[ 0 ]

That's funny, so the "ERROR: " were the actual inconsistency. My next push will have a lot of formatting changes, even if it's not exactly exhaustive because... oh my goodness, this file is... ridiculous

[ 1 ]

now its a font shitting up your browser

[ 1 ]

WTF, sorry i new on github.... some misclick

[ 1 ]

you can use assertin and assertnotin

[ 0 ]

iirc dumb mapping of non dumb buffers is not supported it happens to work on some drivers but we should probably not rely on it too much

[ 0 ]

this is such an irrefutably stupid, awful, and just downright bad idea but it's so fucking funny in theory that it just might work...

[ 1 ]

Yea, what i originally put in but the sprite was fucky when testing, prob how the old file was formatted so I'll change it

[ 1 ]

Why are there so many issues with angular 8? Its getting ridiculous

[ 1 ]

JFC @cpojer fix jest. This is ridiculous. @spicyj - did you by chance test with iojs+jest@0.5 first?

[ 1 ]

Damn, we need to support switching global css asap :)

[ 1 ]

This sound still makes me physically cringe whenever I hear it

[ 1 ]

seems I somehow screwed it up :see_no_evil:

[ 1 ]

Hey, Hugh, I'm not wild about the line length changes that autopep8 is making. The line breaks are pretty clumsy, and I think they make the code harder to read in many cases. Can you turn particular change off?

[ 1 ]

the ridiculous highlighted error line is here, since it seems highlighting changes after job completition (I assume due to pod logs vs the uploaded logs?) https://gist.github.com/BenTheElder/691ec52da5641df109374b5ba133febd

[ 1 ]

This is really bizarre. What happens when you load https://keybase.io in a browser?

[ 1 ]

Damn fancy acronyms, who did that?

[ 1 ]

stack sh is run under a mere user so i think yes

[ 0 ]

i meant fk with on delete cascade

[ 0 ]

Goofball doesn't want the normies to have his cool tools.

[ 1 ]

i think the part about the user is responsible can be removed block on l will be executed anyway as i read it

[ 0 ]

need tenants more api keystone tenant list request domain none paginate false

[ 0 ]

i think we can kill this using jsonpath rw ext

[ 0 ]

this seems to me to be another id id problem did we ever decide the proper usage of id in the context of cloudcafe

[ 0 ]

uh i am assuming these defaults are really just dummy values you should probably remove them

[ 0 ]

this init is useless

[ 0 ]

this exec should be removed we do not need to minify

[ 0 ]

> making blob weaker vs fucking every obstacle FUCKING STOP

[ 1 ]

gay code

[ 1 ]

okay took another pass at this and made the entirety of the promise tracking significantly simpler previous code was copied from webvr path which i did when i was young and stupid a whole years ago

[ 0 ]

does people really find out this hair swag? Because it's fucking disgusting.

[ 1 ]

so this whole check is bunk i only have mocks here because i override the side effect on with one on so this actually fails properly when i check build req destroy called

[ 0 ]

nit i think this should be just another bullet point to join the above list

[ 0 ]

Sorry if this is ugly; kind of hard to initialize a local due to scoping rules.

[ 1 ]

a language that does not have audio plugin of its own seems ridiculous. Still waiting for the flutter team to come with a solution.

[ 1 ]

i think i am missing something here but i do not know this code at all wall so sorry for the dumb questsion if we fallback to the integration bridge i would have expected to see this tried first then if not found attempt to look up the integration bridge

[ 0 ]

you are addict to this construction though it is not a must here i would say it makes reading harder

[ 0 ]

w is a warning that a file should end in a blank newline h is the hacking rule about always having a single char summary of all docstrings something i personally think is silly and since nova also ignores it l i figured we could as well

[ 0 ]

include another test in here (good practice to make sure loops can't go forever even if some dumb shit happens)

[ 1 ]

Dude this is huge. Just edit it and keep the relevant parts. Add only the first few days after genesis. Delete all irrelevant data. Delete all irrelevant html stufff. Do not commit useless stuff in the repo

[ 1 ]

error message should state which module to import the function from, it's ridiculous not to see that on the development console.

[ 1 ]

isn t this what management false is supposed to do i think you can revert this make sure you haven t broken the admin public mechanism

[ 0 ]

yeah i considered that but i also wanted to instantiate this class as a fallback when the configure inspector type cannot be loaded for any reason

[ 0 ]

use die instead of this and the printk

[ 0 ]

An atmos pipe has a height of 1m and a radius of approximately 15cm In freedom units that's a height of about 1 yard and a diameter of about 1 foot You don't want that shit inside of you

[ 1 ]

Er, this could fuck up GCing on ID cards.

[ 1 ]

this needs a docimpact tag please thanks

[ 0 ]

the backslashes were perfectly indented before

[ 0 ]

edit also i am an idiot and it should of course be bytes per pixel in the framebuffer for rgb so same question but make the math and align up

[ 0 ]