00:17 <daurnimator> andrewrk: so how are the 12? :)

00:18 <andrewrk> daurnimator, still on the first one. almost done :)

00:19 <andrewrk> #1733 and #638

00:22 <daurnimator> while on #638, is there an open issue for making readInt take an enum instead of a boolean for the endianess arg?

00:23 <andrewrk> pretty sure it already does. but I'm making a bunch of breaking API changes to all the read and write int functions

00:23 <daurnimator> ah it does! I was looking at an old revision

01:25 <mgxm> andrewrk: https://clbin.com/5mToK

01:25 <mgxm> seems that file_path returned by SplitIterator_rest was not including a null byte

01:26 <andrewrk> ah

01:26 <andrewrk> I'd be happy to make the fix directly, or would you like to make a pull request and get credit for it?

01:26 <andrewrk> nice find, thank you

01:27 <mgxm> no problem, you can fix directly

01:29 <andrewrk> thanks

01:29 <andrewrk> in a moment; I'm in the middle of another commit

02:54 <zachcarter> hello, just stumbled upon this error - https://gist.github.com/zacharycarter/b7bb8ad90306e69399fc0bd97993f3fc

02:54 <zachcarter> I'm sure I'm doing something wrong with the defer - since I'm catching errors already

02:59 <andrewrk> zachcarter, what happened is you returned an error from main, so zig is trying to print an error return trace, to show where the error came from (although in this case it is trivially obvious) and the debug info code itself has panicked and triggered a stack trace, which then succeeds and shows where in zig's debug info code it failed

03:00 <andrewrk> so it's a bug in zig standard library with regards to macos stack traces

03:01 <zachcarter> gotcha - just wanted to make sure that it was already known about, and if not to raise it :) thanks for the explanation

03:01 <zachcarter> enjoying zig so far :) coming from C++ and Nim

03:03 <andrewrk> zachcarter, I filed an issue on your behalf: https://github.com/ziglang/zig/issues/1827

03:03 <andrewrk> thanks for the report!

03:04 <zachcarter> sure thing!

03:05 <andrewrk> mgxm, I created a temporary freebsd vm and I'll see if I can fix that issue

03:38 <andrewrk> mgxm, fixed it.

04:47 <andrewrk> funny, this is almost exactly what I'm doing in copy-elision branch: https://blog.tartanllama.xyz/guaranteed-copy-elision/

05:59 <zachcarter> I'm very impressed with Zig

05:59 <zachcarter> managed to get a basic SDL2 app working with knowing almost nothing about the language in about an hour

06:00 <zachcarter> and that included looking up syntax I didn't understand

06:00 <andrewrk> zachcarter, glad to hear it

06:00 <andrewrk> I'm guessing you ran into this issue along the way: https://github.com/ziglang/zig/issues/1059

06:00 <zachcarter> I very much like many of the design decisions you've taken with the language

06:01 <zachcarter> I did - it threw me for a bit

06:01 <zachcarter> then I found some SDL2 wrapper code and then finally I found @ptrCast and all was right with the world

06:02 <zachcarter> Nim handles this with unchecked arrays

06:04 <zachcarter> I like the proposal though - it's not an easy problem to solve IMO

06:04 <zachcarter> heh - considering this is even an issue - kind of makes me really happy

06:05 <zachcarter> amazed how few bugs I ran into given how new the language is

06:10 <andrewrk> I'm glad to hear you could pick it up quickly though. that's one of my goals - is that source code is discoverable and readable even to people who don't know the language

06:10 <andrewrk> zig code is intended to be very boring and obvious

06:15 <zachcarter> I am a fan of simplicity

06:15 <zachcarter> I like C++ - I think it has some great features

06:15 <zachcarter> when I code in it - I stick to a minimal subset if possible

06:17 <zachcarter> but I'd rather code in a language that is more in the vein of C when possible

09:34 <mgxm> andrewrk: great, thank you

22:30 <daurnimator> would KernelAllocator be a better name for DirectAllocator ?

22:31 <daurnimator> Also other thing I've been thinking about, do we have a way to add misc attributes to memory and/or memory allocators?

22:31 <daurnimator> e.g. traits on an allocator that say "executable"

22:33 <emekoi> it should be possible

22:34 <emekoi> here's an example of how that would work in c https://nullprogram.com/blog/2015/03/19/

22:42 <daurnimator> emekoi: I know how it works in C..... the question is how we can apply attributes to memory and/or allocators in zig. It's more than executuable too: not all memory is the same. e.g. on some architectures, some memory is faster to access than others. you might decide to have 'persistant' memory. huge pages exist...

22:45 <zachcarter> got sdl2 + bgfx initialized with zig: https://gist.github.com/zacharycarter/9f04383f144761ad01b1791e195bfeac

22:45 <zachcarter> time to start making cubes :)

23:38 <andrewrk> zachcarter, nice!