Datasets:

showvikdbz

/

code-reviews

like 0

useless please remove

this testcase looks useless as you have already covered this l

of course silly me

oh shit is pointblank firing handled elsewhere? Ill take a look at that

dumb question return error here or not currently not seems to me this is okay because the only place that cares at all is the replicatehandoff and that one should continue to work as long as the number of insyncs returned is zero which it is here just wondering if this was considered maybe no need to log here if returning an error

yes since it is browsing our overlord http index page our index page always have the html tag if it is a https server urlopen will not fail it will just return garbage

aah, darn, yes, frustrating...

lint is ok with this why did it choke on im inlinemethod opcode flags data in the original cl then saying that the after is unnecessary

this should probably be testtools testcase testskipped

sdkrootdirectory platform tools adb or even better pathutils join a b c

s nullptr

This var name is balls. Also, it uses 0 and 1 to determine its behaviour despite being undescriptive as fuck.

nit i think an int t is appropriate here and below since size is an int t

shifteroperand may not be able to hold lower bound use addconstant key reg value reg lower bound

i will remove docimpact in the commit message

this sounds scary... not good for raider imo, good for arty

actually looks like there are a bunch you have to run... damn maybe ill just stick with the build

i think clientexception does not have getcode method you should use err http status

WTF. Why did they do that? :confused:

perfect

> all this fucking whitespace this is some kind of joke isnt it

we re killing all enum tests so do not add this one n c

don t you need to use std unique ptr mirror object

on 18.04, NetworkManager is version 1.10.6. So, somewhere between 1.10.6 and 1.22.10 is what I'm looking for. Pretty rubbish range :/ I might have to bisect non-LTS releases until I find the less-broken one.

yeah i spotted this also i will hit it

what was the actual problem see c wip dummy cppunittest with gb cppunittest use library objects vcl for a successful example of using that building it failed on all platforms with the forced cppunit fail in the test demonstrating that linking the test succeded on all platforms remember that you generally need to copy all the gb library settings relevant for linking from the library mk to the cppunittest mk

odd run tests sh p n did not pick that up moved down two lines

i dont think you need to do this M fuckery

need to return here since receiving bad message will terminate the renderer and we should abort processing the message any further

findservicebyname findservicebypid findservicebykeychord foreachservice foreachserviceinclass foreachservicewithflags

this should generate some sort of error the apn database is corrupt

my vote would be to have just one generic flag gdb info is misleading since dwarf can be used by many tools including lldb and libbacktrace i like debugsymbols more i would say it sounds like a dwarfsymbols i e debug sections but not everyone might not know the name dwarf so it is good generic name but both of the above seem to imply that they only control inclusion exclusion of debug data i think it would be conceivable that the flag could have deeper effects say do not deduplicate methods to make sure they can be identified or omit some optimizations that rearrange code anyway this is just bike shedding about names i would not want to change it as part of this cl i picked gdbinformation here somewhat randomly and made sure it is at least consistent with quick and elf writer should i change it back to getincludedebugsymbols and maybe kill gdb in further cl

oh shoot, yeah. i want abs()

6 speedloaders, plus the rounds in the mateba, makes 42 rounds of ammunition for a gun that can 2-tap a guy Holy crap

Your editor sucks :). There is an extra `\\t`.

is this lying looks like you return a build id instead

one issues is that this results in a bunch of logs unnecessarily this is not a big deal but what about just testing if the parsed args id is a valid uuid in which case we treat it as a uuid otherwise we assume it is a name and try to search for it this way there is only one query either way and no uneeded log errors

The duplicate ones? I only removed stuff that appeared twice and shifted the rest in alphabetically. Unless that's intended in which case FUCK

yes this is the only caller that we care about there s another caller that dumps debug info on the host but i have deliberately left that as tl because it is more of a dump of what is stored in the actual headers i did not want this fix in the resource table because it does not care about the frameworks in general this class is supposed to be that glue

Okay, damn I lost this because I unassigned. Taking it back. @neil-marcellini set me straight here that because you clicked from `/concierge`, you should land in `/concierge`, even if it's after signing in.

ngl this being a partial class instead of just another system is really ugly it being in 'nutrition' is also sus but thats fine

getlasterror will be reset right after this call as any winapi call may kill it cache it is value before doing anything else

> > > needs testing to make sure this actually works since its a github webedit lol Gives a shit ton of undefined vars.

How _hell the_ did I make it wrong again? ^^'

remove this useless line please

I have never had so many interdependent changes as this bindings nonsense

i believe the nolint below was for the use of unsigned long here keep it nearby

i think you need a install r tools pip requires

it makes more sense to use trans rather than blocktrans

Nice bug fix! Squash do those bastards while they re down!!!

it is better to add this new target to the list of phony targets phony mirror centos

british spelling pls you dirty yank

mock clients return value environment

this should get removed it is ok for libclcore to have more functions than is necessary and we must always build it with the latest api level rs version

Dang. There's always some dumb thing from a previous design that slips through.

This is literally ridiculous. I deleted the whole Parity and now when I re download it just opens up some stupid DOS window. WHERE THE FUCK DO I SEE MY WALLET ON PARITY???

Yes. That's one thing you were very good at by the way, knowing when to use its versus it's. A lot of native English speakers screw that up all the time.

minor nit maybe change uri to use to trigger to uri used to trigger

should be vmwareapi vmwarevcdriver

@glo-www damn, and when I still fixed this, I couldn't find a better selector, and of course adding a selector wont fix it on every themes since they'll need to update to get the fix and their update wont work anymore :/

oh, damn, you're right. Ugh.

theere is a possibility that the manager would still be running we need to check if any processes left and kill them

hmm okay maybe this makes my network nullable comment moot

If anyone else should find this, I got sshfs working by passing <code>-o allow_other,default_permissions</code> but @aldanor is quite right about it being a no-go since it's just too gad damn sloooooooooow! :unamused:

why not just inline the code here like this if mbisclippathcontent svgtokenmarker mrowner gettype mafill isset

damn.. can't you do this in a different way like process()? This seems like an easy way of doing it, whilst also being really bad

Wow, scratch that previous part, I just reread that line and holy fuck that's stupid. I will fix that.

Alright, it's ugly but it's working. I'll merge it now, refactor can come later. EDIT: Rest of the code looks fine, it's just this one weird pointed bool.

dumb question what is http x service name and when is it used

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

Does it work for non-pre-rendered routes? Could you log from one of them what exactly the incoming request from LinkedIn looks like? The 416 makes me think they are sending some wacky `Range` headers that the static file server we're using doesn't like.

This is a little weird, can the polling component be exposed as a protected reference on the base class rather than doing this local write?

nit subtree extend children

current action nullptr

what the fuck dm give it a space

define s are generally evil use constants instead which are closer to the c style const char kcmdstr cmd

yes i think we should launch a new microversion

yeah slimes are boring as shit to play just remove them from the pool Tee Bee Ech

oh my god, this is terrible. :D not your fault at all but it's terrible

If it\u2019s why the version was changed, because this PR requires a change in the SQL backend If it\u2019s why does it need SQL, save files are fucking awful and barely work

Damn, yeah, it was for the debugging. I'll reverse it.

i thought about it but i hate makefiles and their weird language so i thought extracting the list of files was good enough

I agree. But If i recal this feature request was declined because "its out of the scope of the project" "use something else"... "we already have a solution in place" (and that solution sucks btw).... and many other nonsense reasons. Would be much easier for everyone if ts would translate project absolute paths to relative paths. If the reasoning "many developers need this out of the box" is aint enough.. why is typescript still in development then?

It's a direct copy of eating ice lol. I didn't want to end up killing people - unless they drink a shit ton, then it's their own fault :P

this gzip call is problematic because it removes the gz suffix from the image filename therefore when you run stack sh a second time download image will not detect that the image was already downloaded because it checks for the filename with the gz suffix which was cut off by the gzip call and download it again then gzip will try to decompress this new file which causes a conflict with the existing file

probably something stupid it is restored

its a shitty alternative name for a sprite XD

drop the die

driver number self path regex match disk path

Very atomic shit going on right there

The diff got a lot of rubbish. Basically, the testN() got specific names, and the test suite runs them based on registration (see at the end), so it's easy to add a new test and to run tests based on sub-string match on their names. For instance, you can say "daregress localhost DFS" which will run all DFS\\* tests... @jakesmith please review, no regressions detected.

i missed it but i will use snapshot destroy

should probably put adb shell in front of this since it will likely be cut n paste d

English just sucks...

Face palm! I failed to read the ethereum implementation of `matching_transactions` and also failed to remember that @luckysori had done loads of work and added channels to that implementation. Closing this because with that new information this PR is rubbish.

done use a local object instead this method is not performance critical so allocating a dummy object should not matter much

objects servicelist get by binary calls this and some scheduler host manager stuff calls that but with an elevated context already and cellstatemanager update our capacity also calls that but that always passes an admin context so looks like we are good here

it should be final allowed hosts please also add secure cookies false

4.2.2 and it's still broken... This is becoming a ridiculous joke. Why exactly am I paying money every month to a company that is not able to fix a form validation bug in nearly a year?

the fakeservice class should be used