01:00 <shritesh> andrewrk: Master is broken https://github.com/ziglang/zig/pull/2365 ;)

01:01 <andrewrk> shritesh, oops, thanks

01:01 <andrewrk> I ran the tests and then made a minor change with that breakage

01:03 <shritesh> The tests run much faster with your latest change.

01:03 <andrewrk> the zig fmt thing? that's kind of surprising

01:04 <shritesh> Or maybe the builds are just cached.

01:04 <andrewrk> it moves the compilation of zig fmt from first run to during `make`

01:13 <shritesh> ah

03:15 <emekankurumeh[m]> does zig emit compiler errors if you try to link a library that it can't find? or does it assume that the library will be dynamically loaded?

03:20 <emekankurumeh[m]> also zig doesn't build libc from source on freebsd so does it need to link crtbegin.o and friends?

03:21 <daurnimator> emekankurumeh[m]: yes. on freebsd there is no stable syscall ABI: you must link against libc

03:21 <daurnimator> linux is the only widely in use platform that actually has a stable syscall AB

03:21 <daurnimator> +I

03:22 <emekankurumeh[m]> i thought so but i wasn't sure

03:50 <daurnimator> How can I get an empty error set?

03:52 <daurnimator> I'm trying to do: `const NoErrorOutStream = std.io.OutStream(EmptyErrorSet);`

03:57 <daurnimator> ah! error{}

05:30 <emekankurumeh[m]> have you guys seen: https://github.com/vlang/v and if so what are your thoughts about it?

05:39 <tyler569> I know andrew's fielded questions about it in stream Q&As, he said he won't care about it until it's not vaporware

05:50 <daurnimator> emekankurumeh[m]: that's a very common question. search around for zig and vlang :P

05:51 <daurnimator> emekankurumeh[m]: https://hn.algolia.com/?query=zig%20%22v%22&sort=byPopularity&prefix=false&page=0&dateRange=all&type=comment

06:14 <emekankurumeh[m]> do we have zig benchmarks for things like parser speed or compilation speed?

06:15 <emekankurumeh[m]> preferably for the self-hosted compiler.

06:28 <torque> I have never seen someone tease "soon to release" source code for software before, what an extremely bizarre thing

06:31 <tyler569> it seems to be a bit of a trend in system programming languages these days unfortunately

06:31 <tyler569> v and Jon Blow's jai are in very similar situations

06:38 <torque> I haven't really been following Jai's development but to my knowledge it's more or less a hobby project with no real promises or set schedule. At the very least he hasn't made like 5 different github repos saying "will open source soon" which is I think the weirder part for me

07:14 <daurnimator> torque: I have...

12:52 <Sahnvour> andrewrk, I figured out my issue with linking. my working directory had leftovers from a previous compilation, especially kernel32.lib, which was probably then found by lld. removing it fixed the build

12:53 <Sahnvour> I guess it should avoid looking in the cwd if possible, you never know what's there

15:13 <andrewrk> Sahnvour, aha. I actually have experienced that problem before. I think it might be worth considering how we can prevent that accident

15:14 <andrewrk> I agree we should definitely make it stop looking in the cwd if possible

15:54 <ysgard> How would I do a nested for in Zig? For example, in C I would use "for (int i = 0; i < WIDTH; i++) { for (int j = 0; j < HEIGHT; j++) { ..."

15:54 <ysgard> I thought using a nested while would work, but strangely the outer loop never increments

16:00 <ysgard> Like this seems like it would work: https://gist.github.com/ysgard/1fb5b475d68ec92864f28c7aefccdc95.js

16:01 <ysgard> The documentation describes a for but it looks like it's for iterating over collections

16:23 <tgschultz> ysgard, nested while should work: var i = usize(0); while(i < end):(i += 1) { var j = usize(0); while(j < inner_end):(j += 1) {//do stuff} }

16:24 <tgschultz> your gist won't work because j is never reset to 0

16:25 <tgschultz> so the inner loop only runs once

16:26 <tgschultz> see: https://ziglang.org/documentation/master/#while, specifically the "while loop continue expression"

16:28 <ysgard> Ah, I see! Thanks tgschultz!

18:03 <shritesh> So my zigfmt WASM binary is 837K (197K compressed) in debug mode, 899K (145K compressed) with --release-small, 921K (144K compressed) with --release-safe and 788K (131K compressed) with --release-fast.

19:03 <shritesh> Here's zig fmt running on the web. https://shritesh.github.io/zigfmt-web/ Could be made prettier, and handle errors better.

19:05 <mikdusan> so my browser is doing the format? excellent!