00:03 <Aequus> https://news.ycombinator.com/item?id=17028878 you guys should make a comment about zig

00:15 <MajorLag> andrewrk, huh, I never noticed. When I was doing all the win32 stuff it looked to me like it would just skip anything it couldn't handle.

00:17 <andrewrk> MajorLag, but notice that it did find all the function declarations - even though their types were specified with macro defines

00:18 <MajorLag> yeah, now that you mention it, I should have noticed

00:18 <andrewrk> there's lots of room for improvement, translating macros

16:46 <Heimdall> Hello, i am trying out zig and have a question.

16:46 <Heimdall> How should I format a string?

16:47 <Heimdall> I tried _ = text.appendFormat("x: {}\ny: {}", mousePos.x, mousePos.y);

17:50 <andrewrk> hi Heimdall

17:50 <andrewrk> did it work?

17:52 <andrewrk> oh, I see. I will make an example for you

18:09 <GitHub106> zig/master 6e82107 Andrew Kelley: update std.Buffer API...

18:09 <GitHub106> [zig] andrewrk pushed 1 new commit to master: https://git.io/vpyRE

18:09 <andrewrk> Heimdall, try this: https://github.com/zig-lang/zig/commit/6e821078f625a03eb8b7794c983da0f7793366ab#diff-9f718081e74a3d1a46e8b4e7267313ddR59

18:10 <andrewrk> std.io.OutStream is the struct that gives you all the nice API such as print(), and std.io.BufferOutStream is a struct that gives you an OutStream for a Buffer

19:40 <Heimdall> andrewrk, thank you for the example. It works and helped me :). I got a winapi window that prints the mouse position running.

19:40 <andrewrk> neat!

22:02 <Sahnvour> hi andrewrk, real life caught up on me but I'm still planning to work on windows error stacks eventually, is there any progress on this at the moment ?

22:03 <andrewrk> Sahnvour, not as far as I'm aware

22:03 <andrewrk> windows error stacks would be a very welcome addition

22:03 <andrewrk> especially if we didn't have to depend on msvc being installed

22:05 <Sahnvour> yup. anyways nobody wants to have to use DIA :)

22:06 <andrewrk> even if it only supported debug info that was emitted by ourselves - e.g. LLVM 6 and LLD 6 with the settings zig uses - that would solve 95% of the usefulness of stack traces

22:32 <Sahnvour> I'm wondering if assuming that the .pdb (when present) will be named after the executable and in the same directory is good enough

22:33 <Sahnvour> otherwise, there's .pdb full path inside a COFF section of the executable