<pixelherodev>
My stashed entry now only has to handle references :)
<alexnask>
Yup! I also made all capabilities we dont handle explicitly false
<alexnask>
("referencesProvider":false is in there)
<pixelherodev>
I did that and it didn't work, but I suppose I should try it and see what happens
<alexnask>
Ah, good to know
<pixelherodev>
What's zinput?
<pixelherodev>
I mean, "Zig input" I assume
<pixelherodev>
Ahh, README, yay :)
<alexnask>
We also respond with "not implemented" when we get a method we dont handle but I guess that already didnt work either so... :P
<pixelherodev>
Actualy
<pixelherodev>
That was the problem
<pixelherodev>
I was getting the "not implemented" message as a constant error every freaking millisecond
<pixelherodev>
That's what my patch changes
<alexnask>
Ah, well that sounds like a weird decision on kak-lsp's side
* pixelherodev
shrugs
<pixelherodev>
Alrighty, slaughtered all ZLS processes
<alexnask>
I dont think we can just ignore requests we already explicilty told the client we dont support T_T
<pixelherodev>
Let's see what happens...
<pixelherodev>
Just the opposite
<pixelherodev>
If the client knows you don't support it and asks anyways, don't bother sending an error
<pixelherodev>
Just... ignore it
<pixelherodev>
It means the client is being dumb
<pixelherodev>
Dammit packages still no good :(
<pixelherodev>
But if they worked, everything else new would :)
<alexnask>
Is the zig in PATH on master branch?
<pixelherodev>
... no
<pixelherodev>
That'd do it?
<alexnask>
It should yes
<pixelherodev>
Is there a way to feed it the input separately
<pixelherodev>
?
<alexnask>
wdym?
<pixelherodev>
A way to replace the `zig` invocation with a custom command
<pixelherodev>
Being able to `remotezig` it would be nice :)
<alexnask>
No, let me add an option
<pixelherodev>
For now though, I'll just update the static build
<pixelherodev>
Going to have to kill ZLS, I assume?
<alexnask>
I dont think so
<alexnask>
We just use the zig in path to run `zig run build_runner.zig` next to your build.zig files so just updating zig should do
<pixelherodev>
Dammit
<pixelherodev>
Still doesn't work
<pixelherodev>
Updated Zig, nothing
<pixelherodev>
... never mind
dimenus has quit [Read error: Connection reset by peer]
<pixelherodev>
Just took a bit longer
<pixelherodev>
:D
<alexnask>
It takes a while the first time around yes :)
<pixelherodev>
Really nice work
<pixelherodev>
Hover is awesome too
<alexnask>
It should take < 200 ms if you reopen a file that depends on th same build.zig
<pixelherodev>
That's neat
<pixelherodev>
Plus, I never close it so I'm good anyways
<alexnask>
:-D
<pixelherodev>
Kakoune within tmux has such low overhead that I literally never need to close it
<pixelherodev>
I think I have, like, a dozen open editors at any given momeny
<pixelherodev>
moment*
<pixelherodev>
I'm really impressed, nice work
<alexnask>
Thanks ^_^ More improvements coming soon (tm)
<pixelherodev>
Anyways, going to head in early to make up for lack of sleep last night, hopefully in the morning I'll be able to clean up that stage2 PR :)
<andrewrk>
FireFox317, llvm does not implement the c calling convention for us, sadly. we have to do it ourselves
<andrewrk>
it has a c calling convention attribute, but it does not do everything. we still have to do it anyway
ave_ has quit [Quit: ave on thelounge.lasagna.dev]
lunamn has quit [Quit: ave on thelounge.lasagna.dev]
linuxgemini has quit [Quit: ave on thelounge.lasagna.dev]
<pmwhite>
Is the redisconf talk available yet? Or has that not happened yet?
aerona has quit [Remote host closed the connection]
drp has quit [Read error: Connection reset by peer]
drp has joined #zig
<pixelherodev>
Unrelated: `try jit.reserveI(1).?[0].reserve(3)` I think I like this more, for the API
<pixelherodev>
Instead of returning an error union, the JIT's reserve method returns an optional slice (either the reserved blocks, or nothing)
<pixelherodev>
Reserving instructions *within* the block still returns a potential error though
<pixelherodev>
... no, this is a bad idea; I shouldn't optimize the API for a rare use case
mikdusan has quit [Quit: WeeChat 2.6]
greenfork has joined #zig
drp has quit [Read error: Connection reset by peer]
satchmo has quit [Ping timeout: 272 seconds]
dddddd has quit [Ping timeout: 260 seconds]
drp has joined #zig
ur5us has quit [Ping timeout: 260 seconds]
xackus has quit [Ping timeout: 260 seconds]
tdc has joined #zig
satchmo has joined #zig
stripedpajamas has quit [Quit: sleeping...]
satchmo has quit [Ping timeout: 260 seconds]
cole-h has quit [Quit: Goodbye]
satchmo has joined #zig
mikdusan has joined #zig
satchmo has quit [Ping timeout: 244 seconds]
satchmo has joined #zig
slurpie has quit [Ping timeout: 260 seconds]
satchmo has quit [Ping timeout: 260 seconds]
doublex has quit [Remote host closed the connection]
doublex has joined #zig
satchmo has joined #zig
traviss has quit [Quit: Leaving]
antaoiseach has joined #zig
satchmo has quit [Ping timeout: 246 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 265 seconds]
ur5us has joined #zig
satchmo has joined #zig
lornokto has joined #zig
mikdusan has quit [Ping timeout: 265 seconds]
satchmo has quit [Ping timeout: 244 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 272 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
antaoiseach has left #zig [#zig]
[RMS] has quit [Ping timeout: 240 seconds]
satchmo has joined #zig
lqd_ has joined #zig
[RMS] has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
Snektron has quit [*.net *.split]
D3zmodos has quit [*.net *.split]
ky0ko has quit [*.net *.split]
lqd has quit [*.net *.split]
lqd_ is now known as lqd
satchmo has joined #zig
satchmo has quit [Ping timeout: 272 seconds]
Snektron has joined #zig
satchmo has joined #zig
ky0ko has joined #zig
D3zmodos has joined #zig
satchmo has quit [Ping timeout: 265 seconds]
greenfork has quit [Ping timeout: 256 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 272 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 252 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 272 seconds]
satchmo has joined #zig
satchmo has quit [Ping timeout: 252 seconds]
dermetfan has joined #zig
satchmo has joined #zig
puzzleddev has joined #zig
ur5us has quit [Ping timeout: 260 seconds]
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
alexnask has quit [Quit: Leaving]
satchmo has quit [Ping timeout: 256 seconds]
alexnask has joined #zig
satchmo has joined #zig
lornokto has quit [Read error: Connection reset by peer]
lornokto has joined #zig
satchmo has quit [Ping timeout: 246 seconds]
ifreund has joined #zig
satchmo has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
dingenskirchen has quit [Quit: dingenskirchen]
dingenskirchen1 has joined #zig
dingenskirchen1 is now known as dingenskirchen
satchmo has quit [Ping timeout: 272 seconds]
mahmudov has joined #zig
satchmo has joined #zig
KoljaKube has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
satchmo has joined #zig
alexnask_ has joined #zig
satchmo has quit [Ping timeout: 260 seconds]
alexnask__ has joined #zig
alexnask has quit [Ping timeout: 272 seconds]
alexnask__ is now known as alexnask
satchmo has joined #zig
alexnask_ has quit [Ping timeout: 272 seconds]
st4ll1 has quit [Ping timeout: 246 seconds]
satchmo has quit [Ping timeout: 252 seconds]
mahmudov has quit [Ping timeout: 256 seconds]
satchmo has joined #zig
st4ll1 has joined #zig
satchmo has quit [Ping timeout: 240 seconds]
lornokto has quit [Quit: Leaving]
synaps3 has joined #zig
greenfork has joined #zig
<synaps3>
is zig used in production anywhere ?
<ikskuh>
i hope not
<greenfork>
lol
<ikskuh>
i want to do my hoem automation based on zig
<ikskuh>
but that's all private, extensive testing
mahmudov has joined #zig
<synaps3>
haha, when will it be ready for production ?
<ifreund>
when 1.0 is released :D
doublex has quit [Ping timeout: 260 seconds]
doublex has joined #zig
marijnfs has quit [Quit: leaving]
mahmudov has left #zig ["Leaving"]
_Vi has joined #zig
<fengb>
My emulator is totally production >_>
<ikskuh>
\o/
<ikskuh>
/o\
nycex- has quit [Remote host closed the connection]
nycex has joined #zig
recombinant has joined #zig
recombinant has quit [Client Quit]
layneson has joined #zig
puzzleddev has quit [Ping timeout: 252 seconds]
st4ll1 has quit [Quit: WeeChat 2.8]
layneson has quit [Ping timeout: 246 seconds]
<greenfork>
ikskuh: what home automation are you going to write in zig? something like alarm/locks?
KoljaKube has quit [Ping timeout: 272 seconds]
dddddd has joined #zig
puzzleddev has joined #zig
<ikskuh>
greenfork: i have a completly self-built system here
<ikskuh>
which communicates over serial,tcp,udp and a custom network protocol
<ikskuh>
and i want to write some of my serves in zig
<ikskuh>
they're currently written in rust and i'm not happy about the compilation times
<ikskuh>
and i have to change some stuff anyways
<ikskuh>
so why not use the option to rewrite it in zig
<greenfork>
that sounds nice, a lot of custom parts, I like stuff like this
<ikskuh>
yeah, it's cool :)
<ikskuh>
doens't work right now for most of the parts
<ikskuh>
(only thing that works is clock/weather display)
<ikskuh>
music control is completly broken atm
<ikskuh>
same as deivce power
<greenfork>
sounds familiar :)
<pixelherodev>
andrewrk: ping
dimenus has joined #zig
dimenus has quit [Quit: WeeChat 2.8]
Snetry has joined #zig
slurpie has joined #zig
doublex_ has joined #zig
doublex has quit [Read error: Connection reset by peer]
antaoiseach has joined #zig
antaoiseach has quit [Client Quit]
antaoiseach has joined #zig
<oats>
can anybody here doing zig dev with emacs share with me what their setup is like?
greenfork has quit [Ping timeout: 246 seconds]
<pmwhite>
Sure: "alias emacs=vim"
<pixelherodev>
Boo
<ikskuh>
alias emacs=ed
<ikskuh>
alias vi=ed
<pixelherodev>
alias vi=kakoune
<nephele>
what's with all those sophisticated editors, a needle and a steady hand.. :)
<pixelherodev>
inb4 butterfly
<BaroqueLarouche>
alias emacs=code
cole-h has joined #zig
satchmo has joined #zig
doublex_ has quit [Read error: Connection reset by peer]
doublex has joined #zig
<pixelherodev>
... okay, I know I'm going to feel stupid in a minute but
<pixelherodev>
What's the correct way to get a [][]cons... oh wait
<pixelherodev>
... dammit
<pixelherodev>
I *was* doing it correctly!
<pixelherodev>
The parameter *type* was wrong!
* pixelherodev
faceplams
<pixelherodev>
facepalms*
<pixelherodev>
It was supposed to take [][]const u8, but was taking []const u8 instead :P
<pixelherodev>
gahhh
<pixelherodev>
`expected type '[][]const u8', found '*const [1][]const u8'`
<ikskuh>
yeah?
<ikskuh>
there's a const missing!
<pixelherodev>
?
<alexnask>
[]const []const u8
<pixelherodev>
Ohhhhh
<pixelherodev>
Yeah
<pixelherodev>
Gah, I hate working with multiple layers of pointers :(
<pixelherodev>
I always make some stupid mistake here
antaoiseach has quit [Ping timeout: 264 seconds]
antaoiseach has joined #zig
<fengb>
const * const int const
xackus has joined #zig
waleee-cl has joined #zig
<pixelherodev>
I think we need a way of splitting up tests from within the running test
<pixelherodev>
`./zig build test-stage2` shows one test, for instance
<pixelherodev>
But in reality, there can be dozens of cases all utilizing the shared harness
<pixelherodev>
Maybe a builtin, or something?
<andrewrk>
pixelherodev, you're trying to only run 1 test? is that what you're saying
<companion_cube>
hmmm I could watch the "zig for writing a redis module" talk, but the async/await talk was the, hum, short version 🙃
<companion_cube>
is the long version around?
<pixelherodev>
andrewrk: no
<pixelherodev>
I'm saying self-hosted tests show as a single test
<pixelherodev>
`test 1/1: self-hosted`
<pixelherodev>
It would make more sense for each test case to be distinct
<pixelherodev>
but there's no good way to do that currently, as far as I know
<pixelherodev>
andrewrk: on a different note, figured out a solution to the contributor-friendliness issue with 5422 :)
<pixelherodev>
You give it strings as you showed (`line:column:msg`), it extracts the three components, calculates the byte offset in the source, and generates an ErrorMsg
<pixelherodev>
Then, all three error types are compared against the single array
<andrewrk>
pixelherodev, interesting. my instinct would tell me to do it the other way - convert the actual error message to line:column just like it already does when printing errors to the console, then compare that
<andrewrk>
the code to go in the reverse direction doesn't need to exist
<pixelherodev>
Either way, it's about the same amount of work
<pixelherodev>
The biggest difference is when it happens; this way, it only happens once per expected error message
<pixelherodev>
Otherwise, it'd happen once per received error message
<pixelherodev>
At the end of the day, I don't think it's worth bikeshedding the exact implementation of something so trivial :P
<pixelherodev>
If this is the biggest problem you find with the PR, I'll be happy :)
layneson has joined #zig
wozeparrot has joined #zig
<andrewrk>
yeah, thanks for doing that. I'll be happy if it's easy to add new test cases :)
<andrewrk>
(and update existing ones when they break)
opDispatch is now known as B4s1l3
B4s1l3 is now known as opDispatch
<pixelherodev>
That's the goal :)
<pixelherodev>
Oh, and while you're here
<pixelherodev>
The reason the coercion is needed is *internal*
<pixelherodev>
That is, it's reading the call target's name as an [N:0]u8 instead of a slice
<pixelherodev>
I think that's a deeper bug, so I'm going to need to remove the coercion and investigate before merging
<pixelherodev>
When the call target exists, the coercion is unneeded (which implies it's read as a slice)
<andrewrk>
why is it reading a name instead of referencing a @foo or %foo ?
<andrewrk>
I don't fully understand the problem but the solution is certainly not to break the rules of type coercion
layneson has quit [Ping timeout: 272 seconds]
<pixelherodev>
Right, I know
<pixelherodev>
I'd assumed the coercion was *internal*
<companion_cube>
ah, found the actual async/await talk URL in the logs, nvm
<pixelherodev>
Which was a faulty assumption, clearly
<pixelherodev>
It *is* referencing @foo
<pixelherodev>
@notafunc
<pixelherodev>
The name of the function is being read in by analyseInstCall
<pixelherodev>
It's given type [8:0]u8
<pixelherodev>
Then it tries coercing to a u8 slice, which fails/failed
antaoiseach has left #zig [#zig]
stripedpajamas has joined #zig
<andrewrk>
pixelherodev, that's the actual type coercion function of zig! :D
<andrewrk>
after I finish up this vaddr allocation scheme I'll take a look at the example you're talking about
<pixelherodev>
Nah, I'll get it :)
<pixelherodev>
I've already figured out most of the code paths involved
<pixelherodev>
Alrighty, error cases working again :)
<andrewrk>
nice
<pixelherodev>
... okay, this is weird. After reverting the coercion, I don't see the "failed coercion" error anymore, I just get "missing expected error"
<pixelherodev>
Error trace is actually completely empty...
<pixelherodev>
Back to gdb...
<pixelherodev>
Ahh, okay; this is weird
<pixelherodev>
`call(%notafunc)` errors even when the containing function isn't referenced
<pixelherodev>
`call(@notafunc)` does not
<pixelherodev>
andrewrk: which of those is correct?
casaca has quit [Quit: leaving]
casaca has joined #zig
dongcarl has quit [Read error: Connection reset by peer]
<pixelherodev>
If the func is exported, *then* the coercion error appears
<pixelherodev>
The problem is in resolve_const_string, I believe
decentpenguin has joined #zig
gpanders has quit [Ping timeout: 256 seconds]
gpanders has joined #zig
<pixelherodev>
Hmm, do `@foo` and `%foo` get resolved differently by analyzeInst?
gpanders has quit [Ping timeout: 264 seconds]
gpanders has joined #zig
gpanders has quit [Read error: Connection reset by peer]
gpanders has joined #zig
mikdusan has joined #zig
<pmwhite>
ArrayList insert is segfaulting if I try to insert at index 8 when it has length 8. is the `self.ensureCapacity` call not doing its job?
<pixelherodev>
pmwhite: no, that's correct
<pixelherodev>
If length is 8, the valid indices are from 0..7
<pmwhite>
but `ArrayList.insert` calls `self.ensureCapacity(self.items.len + 1);`, so the length becomes 9.
<pmwhite>
Using a `u8` instead of `[2]f32` is fine, so I bet it has something to do with this being a pointer instead of an array.
cole-h has quit [Quit: Goodbye]
cole-h has joined #zig
<pixelherodev>
Ahhhhh
<pixelherodev>
If the declref doesn't exist, it's read in as "bytes" instead of as a "declref"
<pixelherodev>
which then fails to coerce
<pixelherodev>
Rather, the name is *always* going to be `bytes`, but for existing declrefs, it never needs the name as a constString so it has no effect
doublex has quit [Read error: Connection reset by peer]
doublex_ has joined #zig
<dermetfan>
Sorry, pedantic question: When I pass a struct somewhere which memory the calling code owns, would you consider it a better practice to take a pointer explicitely? `fn init(ref: X)` vs `fn init(ref: *X)`
<pixelherodev>
No
<pixelherodev>
Unless you *need* to modify the original, you should take a raw structure
<pixelherodev>
The compiler will convert it to a pointer if it so chooses
<pixelherodev>
If you want to modify the original, you *have* to take a pointer though
<dermetfan>
Thanks, that sounds good. Will do :)
<pixelherodev>
No problem! :D
<pixelherodev>
I think Azure is broken
<pixelherodev>
It seems to complete successfully, report that it failed, and then just wipe the logs
metaleap has joined #zig
<pixelherodev>
andrewrk: 5422 is good to go
<pixelherodev>
Unless there's anything else you want me to fix
<pixelherodev>
Keeping an eye on Azure CI, to see if it's actually passing or not
satchmo has quit [Ping timeout: 260 seconds]
companion_cube has quit [Quit: WeeChat 2.3]
doublex has joined #zig
<pixelherodev>
Hmm
doublex_ has quit [Ping timeout: 264 seconds]
<pixelherodev>
stage2 incremental compiler thoughts: what happens if you want to change target midway through? e.g. change musl versions, or libcs?
<pixelherodev>
need to restart?
<pixelherodev>
At present, I assume that's the case
<pixelherodev>
Would be interesting to have a mode to compile for various targets, preserving the majority of the ZIR analysis
doublex has quit [Ping timeout: 260 seconds]
companion_cube has joined #zig
stripedpajamas has quit [Quit: sleeping...]
slurpie has quit [Ping timeout: 264 seconds]
satchmo has joined #zig
dermetfan has quit [Ping timeout: 252 seconds]
satchmo has quit [Ping timeout: 272 seconds]
cole-h has quit [Quit: Goodbye]
cole-h has joined #zig
wootehfoot has quit [Read error: Connection reset by peer]
wootehfoot has joined #zig
doublex has joined #zig
satchmo has joined #zig
<satchmo>
totally forgot that the compiler chooses what's best
<satchmo>
one of the nice things in life
stripedpajamas has joined #zig
metaleap has quit [Quit: Leaving]
<pixelherodev>
Something looks wrong with the ggf numbers...
<pixelherodev>
self-hosted with std_gpa looks to be 45% above baseline on the latest commit