#👾-core-development
glass jasper
You will disable dependabot
charred monolithBOT
When the message contains a parameter between $ ($example$) the tag command created will include the parameter as an option. The modal is in case the user wants to set default values for optional params.
limber skiff
@fossil inlet wanna try that branch and see if it works for you?
should have no breaking changes
(as long as you didnt use that one function)
but it should also fix that issue you had
fossil inlet
yea, the imports are fixed
fossil inlet
austere talon
wait why is it empty lmaoo
limber skiff
wait why is it empty lmaoo
weakset
limber skiff
nah I lied partially
limber skiff
@austere talon we should merge this branch before merging dev
austere talon
make pr
limber skiff
alr
limber skiff
yeah I realized it made sense to pass the id too
look at this hack lmao
I could add a length check too, but I'm sure normally it will always have the id as the name
austere talon
wtf
limber skiff
unless someone is pushing something handmade into that
austere talon
do we really need Vencord.Plugins. here
limber skiff
yeah it caused circular import
austere talon
https://stackoverflow.com/questions/17638305/why-is-bind-slower-than-a-closure
Chrome 59 update: As I predicted in the answer below bind is no longer slower with the new optimizing compiler
limber skiff
oh
I mean it doesnt really have a difference because that closure only lasts for the call of toString
but yeah fair now
austere talon
yeah it caused circular import
try to fix that and use direct reference
always accessing via global is not that good
limber skiff
try to fix that and use direct reference
I gotta find how to trigger the problem first
@fossil inlet and that means pinging you ^
fossil inlet
I gotta find how to trigger the problem first
let me try and find a file where it would trigger
limber skiff
got it already
fossil inlet
alr
limber skiff
just gotta import it on a plugin
fossil inlet
i solved it by moving the plugins export to another file, but not sure if thats what you want to do
limber skiff
I dont think it's solvable
fossil inlet
i solved it by moving the plugins export to another file, but not sure if thats ...
whoops, typo
limber skiff
the issue is
Plugin imports patchWebpack -> patchWebpack imports plugins -> plugins tries to get all plugins but they are still initializing
fossil inlet
patches export, not plugins export
i moved the patches export to patchWebpack.ts
and you have to re-export in plugins/index.ts for vesktop
limber skiff
I guess that's the fix
limber skiff
thanks
opaque silo
and you have to re-export in plugins/index.ts for vesktop
buh what does Vesktop need
austere talon
youre welcome
fossil inlet
buh what does Vesktop need
the patches export
charred monolithBOT
charred monolithBOT
austere talon
esbuild update made bundle size a bit smaller
austere talon
you will never guess how this happened
I
- marked
esbuild.build - pressed
Ctrl Shift Lto mark all matches in the file - pressed right arrow to mark one more character (
() (applies to all cursors) - deleted
esbuild.BuildOptions got marked
now I marked one more char so esbuild.BuildO
deleted, so now we have ptions
funny
austere talon
wtf this is trippy https://marketplace.visualstudio.com/items?itemName=Cardinal90.multi-cursor-case-preserve
opaque silo
esbuild update made bundle size a bit smaller
is vencord still bundling 300kb of useless dep into the extensions
austere talon
?
opaque silo
removed rnnoise plugin deps
charred monolithBOT
opaque silo
lmao 😭
austere talon
QuickCss fixed itself and I didn't even change anything
maybe just some bug in old esbuild who knows
limber skiff
before merging we should work on making this use as many promise.all as possible
jagged reef
24 days
austere talon
before merging we should work on making this use as many promise.all as possible
the build script?
charred monolithBOT
limber skiff
the build script?
yeah
austere talon
i don't think it's possible to make it faster lol
opaque silo
vendor....
austere talon
that's buildWeb
who gives a shit
buildWeb is slow but the slowest proccess is just packing extension (which this code is part of)
build is already super fast
300ms
do you really expect it to bundle code that results in 600kb minified js in faster than 300ms
limber skiff
something is slow at starting it
austere talon
wdym
if we split vesktop and desktop into separate things maybe we can make it faster
not really
it saves maybe 40ms
not worth it
austere talon
wdym first start
like first time u run the command?
how to reproduce
can't reproduce
it's always 0.3s for me
show video of what you mean
charred monolithBOT
limber skiff
show video of what you mean
will try next time it happens
rugged spire
wtf this is [trippy](https://github.com/Cardinal90/multi-cursor-case-preserve/ra...
thats so cool
dusk blaze
Hold Up Aint That A Vendor Reference
dusk blaze
@austere talon move somewhere else...
austere talon
<@343383572805058560> move somewhere else...
yop I will move to Denmark
dusk blaze
yop I will move to Denmark
granite sky
I still have no idea where they are
fossil inlet
I still have no idea where they are
are you on vesktop or discord desktop?
granite sky
Vesktop
charred monolithBOT
fossil inlet
Vesktop
run it from console and see if there are any errors relevant
if not, try deleting the extension cache
granite sky
[4244:0211/182903.633686:ERROR:CONSOLE(1)] "Request Autofill.setAddresses failed. {"code":-32601,"message":"'Autofill.setAddresses' wasn't found"}", source: devtools://devtools/bundled/core/protocol_client/protocol_client.js (1)
huh
that's weird
welp
granite sky
if not, try deleting the extension cache
rm -rf .config/Vesktop/ExntensionCache/* - no changes at all after restart
fossil inlet
[4244:0211/182903.633686:ERROR:CONSOLE(1)] "Request Autofill.setAddresses failed...
nah, i get that too
fossil inlet
flatpak horror
granite sky
Okay, i got react dev tools on flatpak version
granite sky
- HOW TF IT GOT disabled
- I thought
killall vesktop && vesktopcounts as a full restart
Okay, after a second restart i got them
thank you vee
But i still wonder how it got disabled...
granite sky
Screw you
austere talon
incoming bump pr? 
leaden furnace
merge 
Pull Request #2780
charred monolithBOT
scenic brook
Who next
left sequoia
merge <:fr:1024751426750132284> [`Pull Request #2780`](<https://github.com/Vendi...
PR approval waiting room gang
i want pwa shit too D;
scenic brook
I have a handful of PRs at this point but I'm not really in a rush about any of them so 
austere talon
@limber skiff let's merge https://github.com/Vendicated/Vencord/pull/3217 and https://github.com/Vendicated/Vencord/pull/3218
austere talon
should we omit empty fields in reporter webhook?
limber skiff
ya
austere talon
jolly egret
would shrink it a lot
odd heath
Yeah we should get rid of it
opaque silo
no breaks just include a passed line in the type embed
limber skiff
@austere talon I guess in another pr we can make reporter not include the embeds for clean things and also move patch timings from bad patches
and in another one I can try to cook up caching finds
also what's this ts-expect-error for?
works fine without it for me
austere talon
also what's this ts-expect-error for?
cause I removed checkJs from tsconfig again
I added it for testing
but it's kinda annoying so I turned it off again
limber skiff
I see
austere talon
it flagged some other js files i didnt wanna change
we can add @ts-check to the build scripts
to have them be checked
austere talon
but it's kinda annoying so I turned it off again
it was giving so many type errors on this
yeah lemme add ts-check comments
see
limber skiff
but honestly you used that ts-node package which is so fucking slow at compiling
austere talon
yeah tsx better
limber skiff
because vesktop has its own preload
I just thought it was wrong
limber skiff
yeah tsx better
yeah surely
charred monolithBOT
austere talon
nah it's easy
limber skiff
bet
austere talon
do u really wanna switch to ts
limber skiff
do u really wanna switch to ts
yeah
limber skiff
the time I complain about isn't milliseconds
austere talon
it's half a second slower
limber skiff
in my computer the first run takes like 3 seconds or something to get finished
I dont know why
austere talon
in my computer the first run takes like 3 seconds or something to get finished
windows problem probably
limber skiff
maybe
austere talon
you still get ts
it literally has no difference besides the fact that you need to type with comments
even here you get autocomplete
typescript in js got really good now
limber skiff
I feel we should have something like this:
const defines = mapToString({
IS_STANDALONE,
IS_DEV,
IS_REPORTER,
IS_UPDATER_DISABLED,
IS_WEB: false,
IS_EXTENSION: false,
VERSION,
BUILD_TIMESTAMP
});
eh i didnt care enough
limber skiff
do you want me to do it
austere talon
i thought about it but decided it's faster to just manually do it xD
limber skiff
funny
a tool like that which uses esbuild under the hood should just be finished and have nothing more to do
they dont even handle the parsing of typescript lol
oh yeah look
limber skiff
also I did run bun watch and then -> pnpm build
austere talon
are u using git bash??
limber skiff
ehh
I cant test now
it's fast in both git bash and powershell now
gotta wait the build files to cool down 🔥
fossil inlet
@dusk blaze i love accidently making reload loop
fossil inlet
isnt that just indirect eval
Everything else, including invoking it via an aliased variable, via a member access or other expression, or through the optional chaining ?. operator, is indirect.
fossil inlet
||i only know that because i read MDN in my freetime ||
fossil inlet
bundler could save so much bytes if they changed to that
wont that not throw if x is undefined
austere talon
(0, x)() => x?.()
but this is not technically correct
because it changes behaviour
bundlers want to keep behaviour 100% consistent
brazen bone
They could also save quite a lot of bytes if they didn't compile away foo()?.bar() into (x = foo()) !== undefined && x !== null ? x.bar() : null
austere talon
imagine you have code like
try {
return someFunctionThatMightNotExist();
} catch {
return fallbackFunction();
}
your bundler would make it return undefined instead of calling the fallback
fossil inlet
They could also save quite a lot of bytes if they didn't compile away `foo()?.ba...
isnt that compiling for older targets?
brazen bone
Yes
limber skiff
They could also save quite a lot of bytes if they didn't compile away `foo()?.ba...
actually it's not like that 
I dont remember actually but they do use != null
austere talon
should reporter post a message here whenever something new breaks?
just a short message
nimble pendantBOT
Something just broke on 
Canary. See #1337479880849362994
austere talon
so we don't have to manually check #1337479816240431115
charred monolithBOT
scenic brook
Yeah that's kinda what I was suggesting here
Just anything to draw attention to it
Title change idea is bad though because you can only change the channel title a couple of times within ten minutes and they push more frequently than that every so often
charred monolithBOT
fossil inlet
@odd heath dont you have a pr for this already
gritty iris
am I blind
odd heath
Yeah
gritty iris
love being blind
odd heath
odd heath
Something just broke on <:canary:1252721945699549327> Canary. See <#133747988084...
Make it ping devs 
glass jasper
Make it ping devs <:trolley:1024751352028602449>
Better yet make it everyone
gritty iris
real
dusk blaze
Better yet make it everyone
hi @everyone vencord broke lol
fossil inlet
iirc you have to do VERSION with JSON.stringify
limber skiff
oh is it already string
fossil inlet
oh is it already string
yea, but esbuild cant parse it as anything
idle fossil
I saw an issue from a while ago about arRPC causing significant CPU usage via horrendous WMI queries but it was closed as upstream, are the people who maintain arRPC around here?
just spent a WHILE trying to figure out what was slamming wmi and turns out it's vencord sob
fossil inlet
I saw an issue from a while ago about arRPC causing significant CPU usage via ho...
@mortal forum is the creator of arrpc iirc
idle fossil
o oka
i also see a pr on arrpc that replaces the entire win32 logic
with windows api stuff instead of wmic
looks kinda horrible and unmaintainable if anything ever breaks with it though lmao
austere talon
lmao why u overcomplicating
limber skiff
not really
limber skiff
why tho...
austere talon
I used ignore for a reason 
expect-error is so buggy
it randomly starts saying "there is no error remove this expect error"
then you remove it and there is actually an error
limber skiff
yeah but these
austere talon
overcomplicated and then you have to overcomplicate stringifyValues and then stringifyValues doesnt make a lot of sense cause you have to manually stringify strings
austere talon
just leave it the way it is now 😭
limber skiff
inconsistency tho
austere talon
huh why did u reorder these
austere talon
tsc is also slow and imo least useful
well doesnt matter
i never run test command
thats only for ci
charred monolithBOT
austere talon
seems to work fine
austere talon
true lol
limber skiff
just change those two too
limber skiff
lmao
charred monolithBOT
austere talon
surprisingly it still worked
like not even any errors or crashes
but it had some DiscordNative code paths so some plugins would not have worked (like voice messages)
rugged spire
I feel we should have something like this:
```js
const defines = mapToString({
...
i love Object.entries
scenic brook
3/10 music not angry enough
granite sky
too expensive
scenic brook
$3.50 final offer
charred monolithBOT
rugged spire
median rapids
common code berg L...
shell shuttle
because QA
already tried it
works on my machine up until i removed a single period in the description and now the buttons show on emojis 😭
idk if it's just me or an everyone thing
wait i might have removed needed code 
austere talon
common code berg L...
not even an L...
same thing happened to me on github
spammers exist everywhere
median rapids
never been spammed on gh
limber skiff
@austere talon time to merge both prs
austere talon
ye
limber skiff
nvm
I changed target to dev
and now there's a conflict
lmaoo
right, because my branch touched that config too
limber skiff
I'll resolve the conflict for you and merge
limber skiff
vtest dev
limber skiff
It's caused by Vencord
I guess our update completely bricked windows installation
for when you need to fully open the APP
(which I only tested by reloading)
austere talon
zombie process?
that error means you have two instances
there's probably a zombie process
kill with task manager
limber skiff
I did
it's caused by Vencord
is it somehow opening two instances?
completely diferent ouput for when it works
reverting the last commit fixes it
austere talon
so this?
limber skiff
yeah
limber skiff
let me build without minify
limber skiff
LMAOO
austere talon
you did that in https://github.com/Vendicated/Vencord/pull/3218/commits/1304507f17ad5dbe93290ccf1aa8f51f732b6099
austere talon
you made it bundle like no code in main
ofc it won't load
it never executes discord code
charred monolithBOT
limber skiff
alright
limber skiff
I'm gonna merge this better folders fix into main
and then do merge commit in dev
(who cares)
austere talon
sounds good
discord app works for me
Oh my god I have all plugins enabled on Discord Desktop cause of testing
limber skiff
still good for us to keep it in dev for some more days
austere talon
it's so unusable
have to press ctrl enter to send messages 😭 and poggersmode and oneko
limber skiff
Oh my god I have all plugins enabled on Discord Desktop cause of testing
did you not use reporter desktop 😭
austere talon
nah
reporter makes it take 2 years to load
actually
this was for performance testing
limber skiff
// Decide if we should render the expanded folder background if we are rendering the Better Folders sidebar
austere talon
reporter makes stuff slower
limber skiff
it's not even important
limber skiff
nahh
I fucked up the commit
github is so stupid
changed branch to dev2 -> merged -> realized I needed to make dev2 up to date with main -> force pushed -> merged pr -> pr now includes changes in dev in the commit
limber skiff
vtest dev2
nimble pendantBOT
vtest dev2
Now testing! 
limber skiff
oh
@austere talon does venbot not work for testing branches without the latest changes anymore
austere talon
if ur branch doesnt have the reporter changes no
frail skyBOT
limber skiff
alright
frail skyBOT
austere talon
vtest reporter-no-useless-embed
nimble pendantBOT
vtest reporter-no-useless-embed
Now testing!
limber skiff
should I artifically make one slow trolley
yep
nimble pendantBOT
vtest reporter-no-useless-embed
nimble pendantBOT
vtest reporter-no-useless-embed
charred monolithBOT
austere talon
vtest reporter-no-useless-embed
nimble pendantBOT
vtest reporter-no-useless-embed
Now testing!
limber skiff
nahh wtf is that 😭
nimble pendantBOT
vtest reporter-no-useless-embed
nimble pendantBOT
vtest reporter-no-useless-embed
limber skiff
I have it
charred monolithBOT
austere talon
vtest reporter-no-useless-embed stable
nimble pendantBOT
vtest reporter-no-useless-embed stable
Now testing!
limber skiff
true
nimble pendantBOT
vtest reporter-no-useless-embed stable
austere talon
nice
charred monolithBOT
austere talon
vtest reporter-no-useless-embed stable
nimble pendantBOT
vtest reporter-no-useless-embed stable
Now testing!
nimble pendantBOT
vtest reporter-no-useless-embed stable
austere talon
limber skiff
wont that emoji not work in other forks
austere talon
if they use webhooks it will
if they use venbot fork then they are insane
so doesnt matter
limber skiff
lol
austere talon
it still supports webhook
limber skiff
ship it
austere talon
vencord users love reporter commits
charred monolithBOT
limber skiff
bring me that green embed
austere talon
should I enable this
limber skiff
ya
austere talon
I could have sworn it was on lol
It messed up stuff once
cause
WAIT
I KNOW WHY WE TURNED IT OFF
cause if we merge a PR from dev to main, it will delete dev
which closes all PRs that target dev
limber skiff
LMAOOO
limber skiff
do NOT turn that on
austere talon
limber skiff
buddy what
limber skiff
I love when javascript engine just decides they are gonna be slow
austere talon
limber skiff
vtest dev
limber skiff
I'm running it again, poor #1337479816240431115
austere talon
it tests dev by default u dont have to specify
limber skiff
limber skiff
I have a question
should I just try finds caching or should I try to revive my other pr again too
which does a lot of the webpack finding stuff
but this time with caching too
austere talon
oh this is why it's object object
limber skiff
lol
austere talon
oh this is why it's object object
i had to change it to that cause the function also supports both
limber skiff
we can even cache by build -> and then if build changes check every cached id manually and test the find again
austere talon
in which case you need to specify both
limber skiff
if it matches all good, else get the new id and update last build number
austere talon
we can even cache by build -> and then if build changes check every cached id ma...
true
thats probably best
limber skiff
cannot cache patches but can cache their find too
austere talon
only difficulty will be generating key
austere talon
yes but not functions
austere talon
cannot cache patches but can cache their find too
nah we cant
well
with lazy patches we can but then we have to refactor again
limber skiff
why can't we?
austere talon
with webpack finds we dont have time constraint
with patches we can't wait until the right id is required
because if our id is outdated it will never resolve
we have to add listener whenever modules are registered
but even then we can't really rely on it
well we kinda can
limber skiff
factoryListener is triggered whenever they are registered, not required
so it works both ways
eagerly or not
austere talon
- when a module is newly registered, immediately check if our cache has patches for that module id
- when a module is required, we have to check it for all still unresolved patches (because in the case that the patch has an outdated module id we would otherwise skip this module)
we will have to do this
limber skiff
hmmm
I think we need to take a different approach
whenever a module is registered we immediately find patches that will be applied to it
and map that module to them
so, if our cache has an id for a patch find, and the id matches we test the find, if all good keep it, else we test all finds again on that module
but only if the build number changes
austere talon
so, if our cache has an id for a patch find, and the id matches we test the find...
yes exactly what I mean
austere talon
1. when a module is newly registered, immediately check if our cache has patches...
literally what I said here xD
austere talon
whenever module is registered, immediately look up its id in our cache
and then whenever a module is required (loaded) and its id wasn't in cache, we need to test all patches on it
limber skiff
actually
do we need to use factoryListener?
we could do the same when they are required, no?
or am I missing something
austere talon
we could do the same when they are required, no?
yes
i was overcomplicating it in my head
limber skiff
me too
this shit is complicated lol
I think I will do patches first
they are always a string or regex which makes everything easier
and I dont have to touch webpack finds for now which is still a little messy
idle fossil
chat what do we think about changing arRPC to only query processes with open windows
limber skiff
Datastore is async...
are we gonna need to delay discord startup? @austere talon
limber skiff
localStorage then
austere talon
localStorage is kinda shit for many reasons
I would use a file on desktop and localStorage on browser
you can use sync ipc
localStorage might be limited in size
idle fossil
a mere
limber skiff
discord storages much more things than us on it
all blocked domains
(it's 2.5mb)
jolly egret
cause if we merge a PR from dev to main, it will delete dev
incredible
surely that won't happen if you have some branch protection?
or is github that stupid
rugged spire
chat what do we think about changing arRPC to only query processes with open win...
idk how this would even work (doing so would break things anyway) and as such this sounds like spawning new process every time and wmic being depreciated issue
idle fossil
what
limber skiff
oh
idle fossil
what would it break btw?
limber skiff
we could have been using Promise.withResolvers already
fossil inlet
we could have been using Promise.withResolvers already
is that available now?
fossil inlet
idle fossil
its still insane though
why insane? it's better than bombing wmic with queries of every single process
rugged spire
why insane? it's better than bombing wmic with queries of every single process
the querying for processes with open windows part is just ??? to me
do note i might be totally wrong about this as i haven't touched windows (os) stuff like that for a few years
also i'm double stupid this would still break stuff if you filtered it
i bet it is just wmic moment
idle fossil
what
yes querying the system for every running process every 5 seconds is not good
it causes about 20% cpu usage at all times for me on my 7900x
fossil inlet
https://github.com/tc39/proposals/blob/HEAD/finished-proposals.md
i need to read this more
love samsung
limber skiff
why turn something sync into async tho
I'm failing to see how this is useful
hmm
I never really did this
rugged spire
it causes about 20% cpu usage at all times for me on my 7900x
wouldnt you need to query everything and filter to processes with open windows anyway
genuinely just sounds like overhead from wmic or the parser
fossil inlet
i forget why, but i remember needing something like that a while ago
idle fossil
it would probably be similar to the open pr on arRPC
just using open windows instead of all processes because why the hell do we need to do that
for some reason, roblox was open on my pc in the background for 10 hours with no way to access the process and arRPC was picking it up
i think that's a little stupid
rugged spire
just using open windows instead of all processes *because why the hell do we nee...
genuinely i don't think you understand that the open windows thing doesn't matter at all and if anything will break things, for example a detected "app" could have the window open in a different process name
the way it currently works is just wmic process get ProcessID,ExecutablePath /format:csv
rugged spire
because wmic is shit
idle fossil
very helpful
rugged spire
i have a sadly true story from half a decade ago
idle fossil
there should still be an attempt to work around it because it's not gonna change
rugged spire
i had a [bad] laptop back then and i made a script that ran wmic commands on a loop
austere talon
we could have been using Promise.withResolvers already
i wanted to use it in nodejs
but nodejs22
electron is 20
rugged spire
that script made so many garbage processes
and it ate half my ram
because wmic just got stuck
glass jasper
Paid actor
I like i chould affored someone
rugged spire
there should still be an attempt to work around it because it's not gonna change
it should probably be changed
i bet theres still 9999 npm packages that still use wmic
idle fossil
so what lol
rugged spire
semi unrelated but is now a good time to say that you can disable reg.exe on windows with a group policy and discord desktop calls reg.exe directly
rugged spire
so what lol
time to browse NPM to see what other people do instead
idle fossil
i have a feeling most don't need to see every process
rugged spire
oh thats a package on NPM with a vendored EXE to list processes
idle fossil
nice
fossil inlet
*oh thats a package on NPM with a vendored EXE to list processes*
horror
rugged spire
it causes about 20% cpu usage at all times for me on my 7900x
how i feel dropping frames on discord on my laptop
idle fossil
horor
rugged spire
vesktop 1.5.6 now includes 10% more bloat from npm
@idle fossil i am actually curious what your task manager looks like when its doing this
specifically both the performance and details pages
rugged spire
i hope your memory usage isnt due to wmi
rugged spire
i am an insane linux user and i only made my desktop shell leak 9gb of ram one time
genuinely what is taking up 17.8GiB of ram on my computer
idle fossil
stuff
rugged spire
70% of it is just chromium
idle fossil
tbh i just think windows reports it wrong
rugged spire
i am sure that i said a bunch of stupid stuff in this conversation
austere talon
genuinely what is taking up 17.8GiB of ram on my computer
the vencord btc miner: 
idle fossil
*oh thats a package on NPM with a vendored EXE to list processes*
were you lookin at ps-list?
seems like a fine solution
rugged spire
were you lookin at ps-list?
yop
charred monolithBOT
opaque silo
are you guys discussing the insane windows arrpc scanning process
rugged spire
yes
idle fossil
yes
rugged spire
the vencord btc miner: <a:nyanclap:594991180484444201>
vencoin miner 🤑 🤑 🤑
rugged spire
idk probably the fact that it needs an exe
idle fossil
mehh
opaque silo
does ps-list include CLI args
glass jasper
the vencord btc miner: <a:nyanclap:594991180484444201>
was the reddit post taken down 
opaque silo
it just gives you a list of processes (cross platform)
does it include the command line args in its returned list
opaque silo
returns something called "options" so probably?
i dont see that on the npm page
wher
also including all users processes on mac and linux is a bit of an issue...
rugged spire
also including all users processes on mac and linux is a bit of an issue...
who said we have to replace the implementation there
idle fossil
lol real
we could replace all of it but yeah it could just be windows lol
it probably doesn't show processes you run as admin but still
opaque silo
it probably doesn't show processes you run as admin but still
goodbye kernel anti cheat games
limber skiff
@idle fossil rewrite arpc in rust
rugged spire
goodbye kernel anti cheat games
can we make a vencord plugin bans you from vencord as soon as you open a competitive shooter game
limber skiff
peak idea isn't it
opaque silo
<@267706958428110848> rewrite arpc in rust
i was literally doing that to learn rust
idle fossil
supposedly fixes it
limber skiff
native node module
opaque silo
then I got bored
limber skiff
ngl that is cool
idle fossil
idk why they said it's undocumented tho
it is documented
it's just internal and may explode later
idle fossil
OHH the internal ffi code thing is just for querying the specific process' information
otherwise it's just one syscall EnumProcesses
charred monolithBOT
Content
making an issue so it doesnt get lost
Patch Helper no longer shows the correct replacment for $self
this was probably caused by one of the patcher rewrites recently
Request Agreement
- [x] I have read the requirements for opening an issue above
idle fossil
lol
actually after looking at some of the winapi docs their solution seems pretty much fine
glass jasper
if im signed out, then why show errors
you should be killed for even using Grammarly, that thing is a scam
idle fossil
actually after looking at some of the winapi docs their solution seems pretty mu...
only thing is it probably should error after trying once with 0xffff buffer size and failing
otherwise it could go forever
how do i update a pr that hasn't been touched in 7 months and the person isn't in this discord
lmao
idle fossil
what if i fork his fork and pr it back to arrpc
limber skiff
I need to start reading my commit messages before pushing
idle fossil
lol
idle fossil
austere talon
Am I the only one who pushes commit then views their own commit on github to verify it was all correct
limber skiff
and that's why so many times I have to force push again
idle fossil
literally
limber skiff
because I realize the commit message was just WRONG
idle fossil
i wish you could force push on github without anyone knowing 😔
big red force push makes me feel guilty
fossil inlet
ty
austere talon
new github feature that holds back your commit for 5 minutes so you can edit it without force push @limber skiff
scenic brook
Commit QA branch
glass jasper
new github feature that holds back your commit for 5 minutes so you can edit it ...
13 Confirmation prompts asking you if you’re sure you’d like to commit it
idle fossil
chat can i make a pr that's a continuation of another pr
the first pr is definitely dead
wait this is on arrpc no one's gonna see it
opaque silo
ducko appears once in a blue moon and merges
rugged spire
Am I the only one who pushes commit then views their own commit on github to ver...
i will commit, undo commit, recommit and then push
or just merge a broken feature branch instead
glass jasper
i will commit, undo commit, recommit and then push
my website fr
idle fossil
fire emoji
hollow musk
i wish you could force push on github without anyone knowing 😔
just dont push and view it with an editor
glass jasper
I swear all of these key bind plug-ins can just bundled into one thing
fossil inlet
iirc there are at least 2 gif picker plugins
glass jasper
iirc there are at least 2 gif picker plugins
favoritegifsearch and bettergifpicker
jolly egret
Waiter! Waiter! More gif plugins please!!
idle fossil
man i wish i could open multiple gif pickers
shell shuttle
vas is dis
i cant build anymore
i accidentally unpatched venccord
@nuckys
@nucksys
@nyckys
@limber skiff do u know anything about this error 😭
limber skiff
pnpm i
limber skiff
esbuild update
shell shuttle
thgank u ❤️
fossil inlet
fossil inlet
@austere talon is it intentional that your virtual-merge npm package outputs cjs in an mjs file
because it was causing errors with something else i was working on
||fixed it for myself with https://github.com/Vendicated/virtual-merge/commit/e7e3f8ad6e3f96a06e8db1f0b5e2914edccdf3fa, tested and it works with vencord as well ||
idle fossil
nookies
limber skiff
ya?
idle fossil
idk
limber skiff
true
hollow musk
<:husk:1026532993923293184>
just delete it, what's the worst that could happen (i always do it when the error happens)
fossil inlet
just delete it, what's the worst that could happen (i always do it when the erro...
yea, i do the same
hollow musk
deleting index is not something you should do though
fossil inlet
deleting index is not something you should do though
yes and no
it is just a file that git uses to mark that there is an operation in progress
if for whatever reason git exits abruptly (power loss, kill -9, kernel panic, etc...), the file will remain and it is normal to delete it
hollow musk
yes and no
it is just a file that git uses to mark that there is an operation i...
really? i rmb deleting index and it causing git to think that every file was deleted, i just recloned the repo afer that because idk how to fix it
fossil inlet
really? i rmb deleting index and it causing git to think that every file was del...
thats what i think it is, i have no idea if it does anything else
lime stone
and that's why so many times I have to force push again
use git diff HEAD~ HEAD
austere talon
<@343383572805058560> is it intentional that your virtual-merge npm package outp...
yes
if your bundler isnt awful it should work fine
rugged spire
9 replies to ask for a review 😭
too late to say their code is cursed af?
its just css class manipulation and im sure it will break at random
rugged spire
ok i will review that then
dusk blaze
its just css class manipulation and im sure it will break at random
its AI slop
it does work but look at readme
rugged spire
its AI slop
guh
brazen bone
Should be patching the thing that assigns those classes, not doing dom manip
rugged spire
yes
fossil inlet
if your bundler isnt awful it should work fine
i love webpack 
austere talon
i love webpack <:crazy_fucker:922650012419256331>
why are u using webpack 9
austere talon
if your bundler isnt awful it should work fine
the code isn't even used
fossil inlet
why are u using webpack 9
WDYM webpack 9
AFAIK latest version is 5
fossil inlet
the code isn't even used
Yes it is?
austere talon
no
austere talon
WDYM webpack 9
AFAIK latest version is 5
without 9
austere talon
Yes it is?
literally unreachable code
fossil inlet
literally unreachable code
the issue is a bit weird
with webpack, if virtual-merge is concatonated, module will refer to the module it is concatonated into, so it wont be unreachable, and it will overwrite all the exports for that module
austere talon
horror
fossil inlet
horror
its even weirder with rspack
module will only refer to the actual exports if and only if module is used somewhere in the module that virtual-merge is being concatonated into
jolly egret
dom manip
only works in english
literally just css
classic vencord shit plugin PR
austere talon
its even weirder with rspack
`module` will only refer to the actual exports **i...
you will upgrade to v1.0.2
austere talon
>dom manip
>only works in english
>literally just css
classic vencord shit plugi...
be nice 😭
austere talon
maybe we could add a proper plugin for this tho
austere talon
maybe we could add a proper plugin for this tho
used to be against this because it could be easily done with css but it's becoming harder and harder without hardcoding localised messages and the other option :has is laggy now
jolly egret
mm
austere talon
literally unreachable code
lovely code
fossil inlet
literally unreachable code
just looked a bit more into this,
while referencing module in an mjs file is an error with node
when done from a browser script with type=module, it just refers to the global value module
odd heath
maybe we could add a proper plugin for this tho
should I?
odd heath
im just gonna make it methinks
odd heath
why not tho
charred monolithBOT
This plugin lets users customize their chat bar by removing default and plugin-added buttons based on their settings. This enhances flexibility and creates a cleaner, clutter-free UI.
brazen bone
Ah yes, when a pr is closed you just make a new identical one
fossil inlet
literally unreachable code
also its not bundled as unreachable even with esbuild lol
just refers to module as a global var
brazen bone
Okay that typeof x < "u" is pretty clever ngl
left sequoia
Okay that `typeof x < "u"` is pretty clever ngl
yeah some of the minifer logic is mad
austere talon
funny minifier hacks
left sequoia
too bad a lot of it is giga slow
jolly egret
Ah yes, when a pr is closed you just make a new identical one
they closed their own which is the wierdest part
fossil inlet
Okay that `typeof x < "u"` is pretty clever ngl
i think you can only do it with boolean and undefined
austere talon
Okay that `typeof x < "u"` is pretty clever ngl
javascript spec team: we want to add a new typeof value 
later that day: umm actually we can't because it breaks the entire internet
fossil inlet
guhh bigint and undefined
brazen bone
Eh, just rename it so it comes before "u" alphabetically
fossil inlet
not boolean
charred monolithBOT
This plugin lets users customize their chat bar by removing default and plugin-added buttons based on their settings. This enhances flexibility and creates a cleaner, clutter-free UI.
odd heath
BRO
brazen bone
That's what they did with those genes that broke excel after all
charred monolithBOT
