showvikdbz/code-reviews · Datasets at Hugging Face

text stringlengths 1 5.2k	labels listlengths 1 1
fk should be data sources id	[ 0 ]
i mean billion doomed files with the same hash is ridiculous but if we are seriously worried about it we could use an int instead and not worry about overflow that would use less memory and be faster than a guid	[ 0 ]
is there any need to mention that deleted resources count towards the quota limit it is a tribal knowledge and may not be obvious to api users	[ 0 ]
This would be much better if it was scraped and autogenerated rather than having it hardcoded. Updating it every time a hero change is rolled out or a new hero is added is a pain.	[ 1 ]
string returned to user must be translatable wrap it in	[ 0 ]
This fucking spawn though, what does it do???	[ 1 ]
extra indentation on the docstring body entire block should be shifted left four spaces	[ 0 ]
we should have something like this for adb root adb unroot the same race exists there is not there a don t close this socket flag should we be setting that instead	[ 0 ]
useless config file variable	[ 0 ]
why is this default being changed it does seem to be a dumb option to boot why would you enable the api extension then disable the driver lol wut	[ 0 ]
why please explain how "fuck 69" is a quality name for a gun	[ 1 ]
pedant bereaucracy do you have a bug blueprint for this patch	[ 0 ]
based on hacking the zmq import should come after this next section of stdlib imports	[ 0 ]
so in comparing with the other puppet jobs we should create the bundled gems dir and switch to a similar pattern for the installation and running of bundler mkdir bundled gems export gem home pwd bundled gems gem install bundler no rdoc no ri verbose gem home bin bundle install retry gem home bin bundle exec rake lint gem home bin bundle exec rake validate gem home bin bundle exec rake spec also do we want to go back to a template to at least check puppet x and puppet x	[ 0 ]
max limit int cfg conf pagination max limit should include in try block	[ 0 ]
all these is checks smell like you need a visitor you ll also handle issub ismul etc	[ 0 ]
if the host state max over subscription ratio value thin provisioning is true and use thin logic is false we will enter this branch then return false and log error invalid max over subscription ratio this is not we expected how about change to this elif thin provisioning and host state max over subscription ratio	[ 0 ]
hmm i mean if namespace is not supported then command ip netns exec qrouter x cannot work	[ 0 ]
bad merge conflict fix i confess id idn t run pep before i pushed i fell into that trap	[ 0 ]
It is subjective, but this is making me sick.	[ 1 ]
afaics this is b c b b ce e f f f cb d c d	[ 0 ]
these args are not needed and should be removed	[ 0 ]
indentation is messed up here as well, not indented enough	[ 1 ]
I don't even remember the mousedrop shit doesn't work in a sane manner and I was confused myself by that this works.	[ 1 ]
Damn Zuckerberg!! :angry:	[ 1 ]
is not it bad to encourage inlined version callers may already have a cached runtime or class linker and now you are hiding this in helper function	[ 0 ]
typo garbage	[ 0 ]
so it took me writing several comments about how this was wrong to realize it was right if you respin a comment about how any one filter that asserts presence in an empty list obviates all the rest would help slow witted people like me	[ 0 ]
useless v must have version	[ 0 ]
rtt request count k	[ 0 ]
you do not use single collectionhandler methods i think you should just use basehandler	[ 0 ]
You cant code around that? I managed to do it just fine for tricky variables. Better than adding snowflake shit like this	[ 1 ]
leaks a subprocess error when it should be raising devicenotfounderror	[ 0 ]
you can probably remove this last sentence	[ 0 ]
rados py	[ 0 ]
you should specify here your company	[ 0 ]
This doesn't work and intendation is messed up	[ 1 ]
Besides, why would we ever make an API where you HAD to call Cancel before Close? I makes no sense at all and is horrible design. Just put the cancel in the Close() call. It's two separate methods for a reason. It makes no sense and is ugly and a dumb API if you HAVE to call cancel before Close.	[ 1 ]
@Eric-Arellano : I would just do the same thing you did in the description: add an integration test with a(n even more) ridiculous python version, and confirm that you see a relevant snippet from your error message.	[ 1 ]
shebang is sh in the script and x mode it will run from shell but not within the tox ini without sh before it not sure why	[ 0 ]
if ssh rmvdisk fails status should be error deleting	[ 0 ]
Facing the same issue with file explorer particularly. ridiculous that there has been no updates on this.	[ 1 ]
unrelated to this patch just a general comment is this used anywhere if this is actually used it is a bug we should switch to using cros board eclass	[ 0 ]
Guys, you are killing me. This has been in the code base for ages, anything it can do LOG can do as well, it's pure syntactic sugar.	[ 1 ]
god, just put this in the supersoldier file and make it a subtype, dammit	[ 1 ]
out channels	[ 0 ]
and here with tf extractfile as f	[ 0 ]
all the issue i identified and maybe the ones i missed makes me think that this coverage is not enough please consider adding more tests to cover corner cases like the ones i spotted	[ 0 ]
actually let us just keep this no need to kill it off really	[ 0 ]
seems use unretained this should be ok media thread will be destroyed first	[ 0 ]
no i just mean why the hell did you add this literal fetish content	[ 1 ]
default pasword needs to be removed	[ 0 ]
egads emacs rst aspell fail	[ 0 ]
Are these values used anywhere? I think it would be simpler to move this functionality to a proc that doesn't fuck with the icon.	[ 1 ]
is it ok to have dead pbr	[ 0 ]
i think first valid should be removed	[ 0 ]
damn , not letting people pick up closets full of people so they can unleash a minion army of excel against IH ? lame.	[ 1 ]
unmanage is implemented for all drivers in base class as empty one i just remove data was was stored for share that anyway will be unregistered so i remove garbage here	[ 0 ]
Age-Gay(nah, nevermind). Forget it.	[ 1 ]
fake clients fakes fakeclients fake clients keystone fakes fakeclient fake clients keystone services create mock magicmock return value service	[ 0 ]
oops this must have been a bad merge conflict resolution	[ 0 ]
nit iterables getonlyelement	[ 0 ]
@dakcarto @m-kuhn can this be closed in favour of #3964 ? This PR queue is getting ridiculous :)	[ 1 ]
added a comment	[ 0 ]
i see red eradicate deredify	[ 0 ]
nit endif config usb pd tcpm mux	[ 0 ]
if you rebase on my debug patch this should go away	[ 0 ]
yup agree maybe we can reduce it to hour for the tests a b g etc so we can complete this test in hours good suggestion i see i will add more comment on top of it the basic idea is that we should not ping dut too often i think at least mins maybe we can choose mins or mins we wanted to test the capability of dut to maintain an idle connection with ap this involves time synchronization wake up from power saving mode dut should remain the connection during the test this recall me that i should also disable power manager powerd to prevent system entering s in wifi firmware it should send out keep alive packets to ap periodically for example every seconds to prevent ap drops it	[ 0 ]
vip management cidr netmask and vip public cidr netmask are useless in this task management and ex network roles are deprecated	[ 0 ]
uint p vendor data	[ 0 ]
what are you some kinda fuckin nerd	[ 1 ]
i think they belong in pathimpl but it is ok to leave them if geoff objects	[ 0 ]
sure that is why i say it is my mistake i wrote that commented out code but now i see that it is wrong you cannot commit wrong code with the explanation that you have just uncommented it	[ 0 ]
This would totally pave the way for Segways and other tiny, ridiculous vehicles.	[ 1 ]
not important but you left a coma here	[ 0 ]
potentially stupid question here should this be hidden true	[ 0 ]
what you are doing with testaccountmanager could probably be done with less lines of code with a mock class that just mocks revokegaiatokenonserver then you can expect call getlastrevokedrefreshtoken called with with the right argument times i forgot the proper syntax	[ 0 ]
nit for the h dead case errno etimedout probably not that critical since the h dead case is when the device is being kicked anyway	[ 0 ]
@kristastevens Totally, it's very clear they're at risk, but I think there's a difference between "You might screw up your content, but you can always restore an old revision if you need to" and "You might screw up your content, and you'll be out of luck if you do because you don't have Revisions turned on".	[ 1 ]
OH JESUS CHRIST WHAT THE FUCK	[ 1 ]
replace all assert eq with expect eq	[ 0 ]
another newline	[ 0 ]
test fail seems random and wacky on this one.	[ 1 ]
just return instead of . assignment please. This looked fucky at first glance	[ 1 ]
Note: I'm presently working around this with the slightly ridiculous hack of using a regex to remove everything after the `DocumentNode<...>` declaration. This is obviously fragile and not a good long-term solution for us!	[ 1 ]
i guess we can kill this one	[ 0 ]
there is no point in this as it will blow up on the db sync command by catching the exception it is going to have an unhelpful traceback	[ 0 ]
you should initialize user id and user ref to none and then guard your call to get user against bad values	[ 0 ]
ditto use fake instance fake instance obj	[ 0 ]
@tigercat2000 It's weird	[ 1 ]
this is a dependency we should get rid of as soon as possible i know it is just because you ported the tests as is but add it to the cleanup list	[ 0 ]
hell yeah john is right you should not cut cover like that imho	[ 0 ]
great it works	[ 0 ]
edit i must have gotten my difs mixed up i did not see that this was in fact a new line so yeah in that case my comment above is wrong and steve s suggestion was valid so i will instead	[ 0 ]
we could get rid of teardownclass if it only calls super teardownclass	[ 0 ]
self reserved cookies self default cookie	[ 0 ]
and then you can instead do this virt props objects virtproperties from image metadata image meta and get rid of the annoying	[ 0 ]
please also use insecure conf keystone authtoken insecure	[ 0 ]
can we be more specific doesn t std unordered map work here	[ 0 ]
We are actually running into the same issue. A work around which sucks is when you want to run expo go change the runtimeVersion to policy sdk version https://docs.expo.dev/eas-update/runtime-versions/. When you want to send the build to eas you then need to change runtimeVersion to the format eas expects such as 1.0.0 or 2.122 is what im seeing from the docks. Not a great solution so hopefully the expo team has a better way of tackling this.	[ 1 ]

fk should be data sources id

[ 0 ]

i mean billion doomed files with the same hash is ridiculous but if we are seriously worried about it we could use an int instead and not worry about overflow that would use less memory and be faster than a guid

[ 0 ]

is there any need to mention that deleted resources count towards the quota limit it is a tribal knowledge and may not be obvious to api users

[ 0 ]

This would be much better if it was scraped and autogenerated rather than having it hardcoded. Updating it every time a hero change is rolled out or a new hero is added is a pain.

[ 1 ]

string returned to user must be translatable wrap it in

[ 0 ]

This fucking spawn though, what does it do???

[ 1 ]

extra indentation on the docstring body entire block should be shifted left four spaces

[ 0 ]

we should have something like this for adb root adb unroot the same race exists there is not there a don t close this socket flag should we be setting that instead

[ 0 ]

useless config file variable

[ 0 ]

why is this default being changed it does seem to be a dumb option to boot why would you enable the api extension then disable the driver lol wut

[ 0 ]

why please explain how "fuck 69" is a quality name for a gun

[ 1 ]

pedant bereaucracy do you have a bug blueprint for this patch

[ 0 ]

based on hacking the zmq import should come after this next section of stdlib imports

[ 0 ]

so in comparing with the other puppet jobs we should create the bundled gems dir and switch to a similar pattern for the installation and running of bundler mkdir bundled gems export gem home pwd bundled gems gem install bundler no rdoc no ri verbose gem home bin bundle install retry gem home bin bundle exec rake lint gem home bin bundle exec rake validate gem home bin bundle exec rake spec also do we want to go back to a template to at least check puppet x and puppet x

[ 0 ]

max limit int cfg conf pagination max limit should include in try block

[ 0 ]

all these is checks smell like you need a visitor you ll also handle issub ismul etc

[ 0 ]

if the host state max over subscription ratio value thin provisioning is true and use thin logic is false we will enter this branch then return false and log error invalid max over subscription ratio this is not we expected how about change to this elif thin provisioning and host state max over subscription ratio

[ 0 ]

hmm i mean if namespace is not supported then command ip netns exec qrouter x cannot work

[ 0 ]

bad merge conflict fix i confess id idn t run pep before i pushed i fell into that trap

[ 0 ]

It is subjective, but this is making me sick.

[ 1 ]

afaics this is b c b b ce e f f f cb d c d

[ 0 ]

these args are not needed and should be removed

[ 0 ]

indentation is messed up here as well, not indented enough

[ 1 ]

I don't even remember the mousedrop shit doesn't work in a sane manner and I was confused myself by that this works.

[ 1 ]

Damn Zuckerberg!! :angry:

[ 1 ]

is not it bad to encourage inlined version callers may already have a cached runtime or class linker and now you are hiding this in helper function

[ 0 ]

typo garbage

[ 0 ]

so it took me writing several comments about how this was wrong to realize it was right if you respin a comment about how any one filter that asserts presence in an empty list obviates all the rest would help slow witted people like me

[ 0 ]

useless v must have version

[ 0 ]

rtt request count k

[ 0 ]

you do not use single collectionhandler methods i think you should just use basehandler

[ 0 ]

You cant code around that? I managed to do it just fine for tricky variables. Better than adding snowflake shit like this

[ 1 ]

leaks a subprocess error when it should be raising devicenotfounderror

[ 0 ]

you can probably remove this last sentence

[ 0 ]

rados py

[ 0 ]

you should specify here your company

[ 0 ]

This doesn't work and intendation is messed up

[ 1 ]

Besides, why would we ever make an API where you HAD to call Cancel before Close? I makes no sense at all and is horrible design. Just put the cancel in the Close() call. It's two separate methods for a reason. It makes no sense and is ugly and a dumb API if you HAVE to call cancel before Close.

[ 1 ]

@Eric-Arellano : I would just do the same thing you did in the description: add an integration test with a(n even more) ridiculous python version, and confirm that you see a relevant snippet from your error message.

[ 1 ]

shebang is sh in the script and x mode it will run from shell but not within the tox ini without sh before it not sure why

[ 0 ]

if ssh rmvdisk fails status should be error deleting

[ 0 ]

Facing the same issue with file explorer particularly. ridiculous that there has been no updates on this.

[ 1 ]

unrelated to this patch just a general comment is this used anywhere if this is actually used it is a bug we should switch to using cros board eclass

[ 0 ]

Guys, you are killing me. This has been in the code base for ages, anything it can do LOG can do as well, it's pure syntactic sugar.

[ 1 ]

god, just put this in the supersoldier file and make it a subtype, dammit

[ 1 ]

out channels

[ 0 ]

and here with tf extractfile as f

[ 0 ]

all the issue i identified and maybe the ones i missed makes me think that this coverage is not enough please consider adding more tests to cover corner cases like the ones i spotted

[ 0 ]

actually let us just keep this no need to kill it off really

[ 0 ]

seems use unretained this should be ok media thread will be destroyed first

[ 0 ]

no i just mean why the hell did you add this literal fetish content

[ 1 ]

default pasword needs to be removed

[ 0 ]

egads emacs rst aspell fail

[ 0 ]

Are these values used anywhere? I think it would be simpler to move this functionality to a proc that doesn't fuck with the icon.

[ 1 ]

is it ok to have dead pbr

[ 0 ]

i think first valid should be removed

[ 0 ]

damn , not letting people pick up closets full of people so they can unleash a minion army of excel against IH ? lame.

[ 1 ]

unmanage is implemented for all drivers in base class as empty one i just remove data was was stored for share that anyway will be unregistered so i remove garbage here

[ 0 ]

Age-Gay(nah, nevermind). Forget it.

[ 1 ]

fake clients fakes fakeclients fake clients keystone fakes fakeclient fake clients keystone services create mock magicmock return value service

[ 0 ]

oops this must have been a bad merge conflict resolution

[ 0 ]

nit iterables getonlyelement

[ 0 ]

@dakcarto @m-kuhn can this be closed in favour of #3964 ? This PR queue is getting ridiculous :)

[ 1 ]

added a comment

[ 0 ]

i see red eradicate deredify

[ 0 ]

nit endif config usb pd tcpm mux

[ 0 ]

if you rebase on my debug patch this should go away

[ 0 ]

yup agree maybe we can reduce it to hour for the tests a b g etc so we can complete this test in hours good suggestion i see i will add more comment on top of it the basic idea is that we should not ping dut too often i think at least mins maybe we can choose mins or mins we wanted to test the capability of dut to maintain an idle connection with ap this involves time synchronization wake up from power saving mode dut should remain the connection during the test this recall me that i should also disable power manager powerd to prevent system entering s in wifi firmware it should send out keep alive packets to ap periodically for example every seconds to prevent ap drops it

[ 0 ]

vip management cidr netmask and vip public cidr netmask are useless in this task management and ex network roles are deprecated

[ 0 ]

uint p vendor data

[ 0 ]

what are you some kinda fuckin nerd

[ 1 ]

i think they belong in pathimpl but it is ok to leave them if geoff objects

[ 0 ]

sure that is why i say it is my mistake i wrote that commented out code but now i see that it is wrong you cannot commit wrong code with the explanation that you have just uncommented it

[ 0 ]

This would totally pave the way for Segways and other tiny, ridiculous vehicles.

[ 1 ]

not important but you left a coma here

[ 0 ]

potentially stupid question here should this be hidden true

[ 0 ]

what you are doing with testaccountmanager could probably be done with less lines of code with a mock class that just mocks revokegaiatokenonserver then you can expect call getlastrevokedrefreshtoken called with with the right argument times i forgot the proper syntax

[ 0 ]

nit for the h dead case errno etimedout probably not that critical since the h dead case is when the device is being kicked anyway

[ 0 ]

@kristastevens Totally, it's very clear they're at risk, but I think there's a difference between "You might screw up your content, but you can always restore an old revision if you need to" and "You might screw up your content, and you'll be out of luck if you do because you don't have Revisions turned on".

[ 1 ]

OH JESUS CHRIST WHAT THE FUCK

[ 1 ]

replace all assert eq with expect eq

[ 0 ]

another newline

[ 0 ]

test fail seems random and wacky on this one.

[ 1 ]

just return instead of . assignment please. This looked fucky at first glance

[ 1 ]

Note: I'm presently working around this with the slightly ridiculous hack of using a regex to remove everything after the `DocumentNode<...>` declaration. This is obviously fragile and not a good long-term solution for us!

[ 1 ]

i guess we can kill this one

[ 0 ]

there is no point in this as it will blow up on the db sync command by catching the exception it is going to have an unhelpful traceback

[ 0 ]

you should initialize user id and user ref to none and then guard your call to get user against bad values

[ 0 ]

ditto use fake instance fake instance obj

[ 0 ]

@tigercat2000 It's weird

[ 1 ]

this is a dependency we should get rid of as soon as possible i know it is just because you ported the tests as is but add it to the cleanup list

[ 0 ]

hell yeah john is right you should not cut cover like that imho

[ 0 ]

great it works

[ 0 ]

edit i must have gotten my difs mixed up i did not see that this was in fact a new line so yeah in that case my comment above is wrong and steve s suggestion was valid so i will instead

[ 0 ]

we could get rid of teardownclass if it only calls super teardownclass

[ 0 ]

self reserved cookies self default cookie

[ 0 ]

and then you can instead do this virt props objects virtproperties from image metadata image meta and get rid of the annoying

[ 0 ]

please also use insecure conf keystone authtoken insecure

[ 0 ]

can we be more specific doesn t std unordered map work here

[ 0 ]

We are actually running into the same issue. A work around which sucks is when you want to run expo go change the runtimeVersion to policy sdk version https://docs.expo.dev/eas-update/runtime-versions/. When you want to send the build to eas you then need to change runtimeVersion to the format eas expects such as 1.0.0 or 2.122 is what im seeing from the docks. Not a great solution so hopefully the expo team has a better way of tackling this.

[ 1 ]