showvikdbz/code-reviews · Datasets at Hugging Face

text stringlengths 1 5.2k	labels listlengths 1 1
wrap your commit body to like cols	[ 0 ]
we should remove zeromq since it is not an amqp service it is a queue yeah but it is not compliant with the amqp standard	[ 0 ]
from browser s perspective dirty needs ack but from renderer s perspective dirty needs ack	[ 0 ]
delete this c string already get initialized as empty	[ 0 ]
typo in commit message useles useless	[ 0 ]
this creates d out opt i do not think that is what you want	[ 0 ]
d d	[ 0 ]
`particleEmitters`. Again, I realize all other fields are `my*`, but let's keep the new code stink free.	[ 1 ]
stray blank line	[ 0 ]
needs locking against create snapshot because a create snapshot and delete snapshot running at the same time will result in a corrupt snapshot volume	[ 0 ]
Ohh shiiiit, didn't know this worked	[ 1 ]
this class is useless	[ 0 ]
Please try to understand the code you're commenting on.	[ 1 ]
steal the comments from the methods for this and all below	[ 0 ]
nvm above sucks, applied what you had.	[ 1 ]
So I can fix wooden barricades with a welding fuel? wtf	[ 1 ]
run usermod a g kolla haproxy	[ 0 ]
So, emscripten on macOS is basically broken until 1.37.23 is released? That's ridiculous. This should have a hot fix.	[ 1 ]
why not place this in the dirty bits iteration with a comment on why we need it for d d	[ 0 ]
globals yml	[ 0 ]
WHAT THE HELL ARE YOU	[ 1 ]
yes i found the issue when the passwd entry was legitimately missing i was surprised to see both arc networkd processes running as root in ps under normal circumstances the entry should not be missing and the risk of something affecting this daemon is low but security critical code often gets attacked in ways that because something that shouldn t fail to fail e g maybe the bad guy cannot write arbitrary contents into etc passwd but on a normal linux box he might be able to trick a careless cron job into erasing clobbering it is contents through symlink attacks and then subsequent calls to getpwnam fail since chrome os uses dm verity to protect the rootfs contents somebody cannot get the os to return arbitrary data when reading a file but maybe they can get the os to return an i o error if the file data is overwritten with bad content on disk exhaustion attacks can sometimes induce failures that are useful to a hacker e g what if somebody opens up file descriptors then runs a setuid program that fails to open the th fd and reacts in a careless way linux veterans may remember this design flaw in the original posix capabilities scheme	[ 0 ]
here too you must replace bin sh by bin bash	[ 0 ]
please line wrap this to chars	[ 0 ]
i like this patch style you save an useless var	[ 0 ]
nit i should probably change this to just key and secret instead of access key and access secret it s already an access token object after all	[ 0 ]
Since the order does not matter, an `std::set` would make for better readability IMHO because you can then use `count()` rather than that ugly `std::find()`	[ 1 ]
This is never null, is it? And if it is then a coder fucked up. Might want to write something that it is unexpected and/or to report it or something. Although leaving it like this is fine as well	[ 1 ]
Is there a better way to format this hellish macro? A C++11 constexpr function might be clearer, but no C++11 yet.	[ 1 ]
bad indenting here and below	[ 0 ]
ditto redundant but harmless for init since we will have already done this upon selinux android reload policy but required for ueventd or other users of file contexts only	[ 0 ]
whoops, probably, find/replace + shitcode!	[ 1 ]
> BREAKING LINE ENDINGS IN EVERY FUCKING TXT FILE	[ 1 ]
depends on the gc but for us nothing put a dcheck here that resolved null	[ 0 ]
maybe sfx should be added to section sfx li c sfx	[ 0 ]
Looks as though they are now recognized as tables by NVDA, but the keyboard behavior seems a little wacky. I or someone else should double check the ARIA implementation.	[ 1 ]
what are you executing off of rootfs via the shell commandline if your userdebug eng builds just use su shell exec su should cuase you to be in su domain which can do almost anything	[ 0 ]
stupid question why are not we using python compatible print	[ 0 ]
same comment as for connection retry	[ 0 ]
no this will be overwritten every time and since we will kill adbd every time we start the program the setting will be flushed	[ 0 ]
nope don't fucking do this	[ 1 ]
God damn Americans... :gb:	[ 1 ]
it needs two levels of dereference which obviously does not really matter but is kind of ugly imho	[ 0 ]
this sucks. use physiology	[ 1 ]
real nit but can we write this without the id greenthread spawn after self events delayed event uuid id sorry for being pathetic	[ 0 ]
remove pyc	[ 0 ]
I must add I find a 3mbps cap ridiculous...	[ 1 ]
it should though the for each observer macro is dumb and expands to a non const iterator since we only use this internally in this file i think it is ok to leave this as non const	[ 0 ]
maybe say mock ssh channel files	[ 0 ]
hi, is it ridiculous of me to ask when this could get reviewed?	[ 1 ]
i know i should have brought up this earlier but i wonder if we would benefit from moving these into it is own section we could do it in a clean up patch but because it is a time sensitive issue we d need to get it in before havana releases i wonder if it is worth doing it in another patchset in this review	[ 0 ]
This is just a hack with ugly hardcoded stuff	[ 1 ]
try wrapping all of the define in () eg: ```#define blah(x) (... your shit using x)``` The () are superfluous and won't break things, but may cause things to be evaluated correctly.	[ 1 ]
...but this throws a "bad argument definition" compile error when I uncomment COLD. what the hell am I doing wrong?	[ 1 ]
you should probably mark this as wip if you are still working on this	[ 0 ]
gosh darn commit didn't push	[ 1 ]
Hell yeah, no more commented code \U0001f60d	[ 1 ]
i think calling os svc upstart will be a mistake as fedora images will be systemd based should os svc upstart be named os svc daemon or something	[ 0 ]
wacky ass shit (your current implementation is preferred)	[ 1 ]
Also, everything likes it much better when there's a new line at the end of the file. You can even see GitHub bitching about it here :stuck_out_tongue_closed_eyes:	[ 1 ]
why is nulable true needed in the model nullable is not specified and i think the default value of nullable is false	[ 0 ]
Ok I've removed some more wild & wacky handling of this field in favor of consistency.	[ 1 ]
done	[ 0 ]
yes in ndk this was hacked around by manually defining has feature we ll need to fix this when upstreaming to libc i was going to take a stab at this upstreaming on friday	[ 0 ]
basename p then you do not need private cpp extension	[ 0 ]
should we issue a log warning here about junk found	[ 0 ]
WTH is happening on lines 636-642 ?	[ 1 ]
kill the comment	[ 0 ]
we should just delete this entirely and use sys tgkill from linux syscall support	[ 0 ]
assume i do neutron create net foo neutron create port foo get hold of port id neutron trunk create parent port id nova boot nic port id port id the question to answer is this would a trunk without subports function correctly i would think so as a guest i should be able to have my interface connection on my foo network just like any other regular port is it useful probably not as you would expect the the user to go and add subports later on so the other question is do we want to allow for a successful boot off a trunk with at least one subport if so then we should probably fail the binding process rather than hitting the agent	[ 0 ]
with with slaps own wrist	[ 0 ]
this should be bumped to the current proposed microversion v	[ 0 ]
what the hell is going on here	[ 1 ]
please do not do mocks like this mock out the thing you need to mock in the place you need it to be mocked this sort of pattern is what we have with mox based tests in many areas of the code and they are often a disaster	[ 0 ]
could we not do these as get trusts trustor user id myuserid get trusts trustee user id myuserid	[ 0 ]
what if upon trimming all data are deleted and self data is empty	[ 0 ]
silly me that was the initial idea	[ 0 ]
looks like some wacky indentation here.	[ 1 ]
keepalive d	[ 0 ]
please add blocksize and execute args here	[ 0 ]
are these two white spaces after testtools intentional	[ 0 ]
which kernel version is needed because this is a ridiculous bug?	[ 1 ]
I based this on already existing typings for legacy projects (`react-addons-transition-group` and `react-addons-css-transition-group`) and didn't want to be an asshole who removes credits. :)	[ 1 ]
it should kill the service implementation in the browser process as well	[ 0 ]
use the typedef you created in the h mockentryiter	[ 0 ]
But that would have been funny. Adminhelp: GrifferMcGee: "FUCK PLZ DUN BANZ ME"	[ 1 ]
Shit. Want it reverted, or is there a fix?	[ 1 ]
minor oswl version info version info	[ 0 ]
i think it should be op rc ne o op rc ne and condition will make dead code here	[ 0 ]
This sucks... Is the `+1` expected?	[ 1 ]
useless	[ 0 ]
use ssl hash public ssl hash service name service type resource type default value args fail if args length	[ 0 ]
i strongly object against this instead you should use sqlalchemy introspection like this from sqlalchemy import inspection pkg inspect inspection inspect pkg foreign keys pkg inspect relationships	[ 0 ]
What the fuck, Dylan	[ 1 ]
Damn the torpedoes, set it to true	[ 1 ]
you already got environment args looks useless here	[ 0 ]
admin token also below	[ 0 ]
What are you smoking, gargleblaster is probably the most common drink and easy as hell to mix Even the dispensers are easy to reach	[ 1 ]
again it sounds like what the fuck size t is expected here since these are what the fuck vector	[ 0 ]
Next time keep this to a separate branch, and just create branches off of branches if you require one branch's changes. It becomes hell to review if you bundle too many things together.	[ 1 ]

wrap your commit body to like cols

[ 0 ]

we should remove zeromq since it is not an amqp service it is a queue yeah but it is not compliant with the amqp standard

[ 0 ]

from browser s perspective dirty needs ack but from renderer s perspective dirty needs ack

[ 0 ]

delete this c string already get initialized as empty

[ 0 ]

typo in commit message useles useless

[ 0 ]

this creates d out opt i do not think that is what you want

[ 0 ]

d d

[ 0 ]

`particleEmitters`. Again, I realize all other fields are `my*`, but let's keep the new code stink free.

[ 1 ]

stray blank line

[ 0 ]

needs locking against create snapshot because a create snapshot and delete snapshot running at the same time will result in a corrupt snapshot volume

[ 0 ]

Ohh shiiiit, didn't know this worked

[ 1 ]

this class is useless

[ 0 ]

Please try to understand the code you're commenting on.

[ 1 ]

steal the comments from the methods for this and all below

[ 0 ]

nvm above sucks, applied what you had.

[ 1 ]

So I can fix wooden barricades with a welding fuel? wtf

[ 1 ]

run usermod a g kolla haproxy

[ 0 ]

So, emscripten on macOS is basically broken until 1.37.23 is released? That's ridiculous. This should have a hot fix.

[ 1 ]

why not place this in the dirty bits iteration with a comment on why we need it for d d

[ 0 ]

globals yml

[ 0 ]

WHAT THE HELL ARE YOU

[ 1 ]

yes i found the issue when the passwd entry was legitimately missing i was surprised to see both arc networkd processes running as root in ps under normal circumstances the entry should not be missing and the risk of something affecting this daemon is low but security critical code often gets attacked in ways that because something that shouldn t fail to fail e g maybe the bad guy cannot write arbitrary contents into etc passwd but on a normal linux box he might be able to trick a careless cron job into erasing clobbering it is contents through symlink attacks and then subsequent calls to getpwnam fail since chrome os uses dm verity to protect the rootfs contents somebody cannot get the os to return arbitrary data when reading a file but maybe they can get the os to return an i o error if the file data is overwritten with bad content on disk exhaustion attacks can sometimes induce failures that are useful to a hacker e g what if somebody opens up file descriptors then runs a setuid program that fails to open the th fd and reacts in a careless way linux veterans may remember this design flaw in the original posix capabilities scheme

[ 0 ]

here too you must replace bin sh by bin bash

[ 0 ]

please line wrap this to chars

[ 0 ]

i like this patch style you save an useless var

[ 0 ]

nit i should probably change this to just key and secret instead of access key and access secret it s already an access token object after all

[ 0 ]

Since the order does not matter, an `std::set` would make for better readability IMHO because you can then use `count()` rather than that ugly `std::find()`

[ 1 ]

This is never null, is it? And if it is then a coder fucked up. Might want to write something that it is unexpected and/or to report it or something. Although leaving it like this is fine as well

[ 1 ]

Is there a better way to format this hellish macro? A C++11 constexpr function might be clearer, but no C++11 yet.

[ 1 ]

bad indenting here and below

[ 0 ]

ditto redundant but harmless for init since we will have already done this upon selinux android reload policy but required for ueventd or other users of file contexts only

[ 0 ]

whoops, probably, find/replace + shitcode!

[ 1 ]

> BREAKING LINE ENDINGS IN EVERY FUCKING TXT FILE

[ 1 ]

depends on the gc but for us nothing put a dcheck here that resolved null

[ 0 ]

maybe sfx should be added to section sfx li c sfx

[ 0 ]

Looks as though they are now recognized as tables by NVDA, but the keyboard behavior seems a little wacky. I or someone else should double check the ARIA implementation.

[ 1 ]

what are you executing off of rootfs via the shell commandline if your userdebug eng builds just use su shell exec su should cuase you to be in su domain which can do almost anything

[ 0 ]

stupid question why are not we using python compatible print

[ 0 ]

same comment as for connection retry

[ 0 ]

no this will be overwritten every time and since we will kill adbd every time we start the program the setting will be flushed

[ 0 ]

nope don't fucking do this

[ 1 ]

God damn Americans... :gb:

[ 1 ]

it needs two levels of dereference which obviously does not really matter but is kind of ugly imho

[ 0 ]

this sucks. use physiology

[ 1 ]

real nit but can we write this without the id greenthread spawn after self events delayed event uuid id sorry for being pathetic

[ 0 ]

remove pyc

[ 0 ]

I must add I find a 3mbps cap ridiculous...

[ 1 ]

it should though the for each observer macro is dumb and expands to a non const iterator since we only use this internally in this file i think it is ok to leave this as non const

[ 0 ]

maybe say mock ssh channel files

[ 0 ]

hi, is it ridiculous of me to ask when this could get reviewed?

[ 1 ]

i know i should have brought up this earlier but i wonder if we would benefit from moving these into it is own section we could do it in a clean up patch but because it is a time sensitive issue we d need to get it in before havana releases i wonder if it is worth doing it in another patchset in this review

[ 0 ]

This is just a hack with ugly hardcoded stuff

[ 1 ]

try wrapping all of the define in () eg: ```#define blah(x) (... your shit using x)``` The () are superfluous and won't break things, but may cause things to be evaluated correctly.

[ 1 ]

...but this throws a "bad argument definition" compile error when I uncomment COLD. what the hell am I doing wrong?

[ 1 ]

you should probably mark this as wip if you are still working on this

[ 0 ]

gosh darn commit didn't push

[ 1 ]

Hell yeah, no more commented code \U0001f60d

[ 1 ]

i think calling os svc upstart will be a mistake as fedora images will be systemd based should os svc upstart be named os svc daemon or something

[ 0 ]

wacky ass shit (your current implementation is preferred)

[ 1 ]

Also, everything likes it much better when there's a new line at the end of the file. You can even see GitHub bitching about it here :stuck_out_tongue_closed_eyes:

[ 1 ]

why is nulable true needed in the model nullable is not specified and i think the default value of nullable is false

[ 0 ]

Ok I've removed some more wild & wacky handling of this field in favor of consistency.

[ 1 ]

done

[ 0 ]

yes in ndk this was hacked around by manually defining has feature we ll need to fix this when upstreaming to libc i was going to take a stab at this upstreaming on friday

[ 0 ]

basename p then you do not need private cpp extension

[ 0 ]

should we issue a log warning here about junk found

[ 0 ]

WTH is happening on lines 636-642 ?

[ 1 ]

kill the comment

[ 0 ]

we should just delete this entirely and use sys tgkill from linux syscall support

[ 0 ]

assume i do neutron create net foo neutron create port foo get hold of port id neutron trunk create parent port id nova boot nic port id port id the question to answer is this would a trunk without subports function correctly i would think so as a guest i should be able to have my interface connection on my foo network just like any other regular port is it useful probably not as you would expect the the user to go and add subports later on so the other question is do we want to allow for a successful boot off a trunk with at least one subport if so then we should probably fail the binding process rather than hitting the agent

[ 0 ]

with with slaps own wrist

[ 0 ]

this should be bumped to the current proposed microversion v

[ 0 ]

what the hell is going on here

[ 1 ]

please do not do mocks like this mock out the thing you need to mock in the place you need it to be mocked this sort of pattern is what we have with mox based tests in many areas of the code and they are often a disaster

[ 0 ]

could we not do these as get trusts trustor user id myuserid get trusts trustee user id myuserid

[ 0 ]

what if upon trimming all data are deleted and self data is empty

[ 0 ]

silly me that was the initial idea

[ 0 ]

looks like some wacky indentation here.

[ 1 ]

keepalive d

[ 0 ]

please add blocksize and execute args here

[ 0 ]

are these two white spaces after testtools intentional

[ 0 ]

which kernel version is needed because this is a ridiculous bug?

[ 1 ]

I based this on already existing typings for legacy projects (`react-addons-transition-group` and `react-addons-css-transition-group`) and didn't want to be an asshole who removes credits. :)

[ 1 ]

it should kill the service implementation in the browser process as well

[ 0 ]

use the typedef you created in the h mockentryiter

[ 0 ]

But that would have been funny. Adminhelp: GrifferMcGee: "FUCK PLZ DUN BANZ ME"

[ 1 ]

Shit. Want it reverted, or is there a fix?

[ 1 ]

minor oswl version info version info

[ 0 ]

i think it should be op rc ne o op rc ne and condition will make dead code here

[ 0 ]

This sucks... Is the `+1` expected?

[ 1 ]

useless

[ 0 ]

use ssl hash public ssl hash service name service type resource type default value args fail if args length

[ 0 ]

i strongly object against this instead you should use sqlalchemy introspection like this from sqlalchemy import inspection pkg inspect inspection inspect pkg foreign keys pkg inspect relationships

[ 0 ]

What the fuck, Dylan

[ 1 ]

Damn the torpedoes, set it to true

[ 1 ]

you already got environment args looks useless here

[ 0 ]

admin token also below

[ 0 ]

What are you smoking, gargleblaster is probably the most common drink and easy as hell to mix Even the dispensers are easy to reach

[ 1 ]

again it sounds like what the fuck size t is expected here since these are what the fuck vector

[ 0 ]

Next time keep this to a separate branch, and just create branches off of branches if you require one branch's changes. It becomes hell to review if you bundle too many things together.

[ 1 ]