Yusuf Bham

Results 68 comments of Yusuf Bham

Might be worth adding `std.testing.refAllDeclsRecursive(@This())` to uefi.zig to ensure that all the packed struct stuff gets checked and stays fixed fwiw

Ah ok. I assume that would change extern struct stuff then? Because refAllDecls would just be to ensure it stays compiling, not much else afaik.

Tested this with my bootloader and it works again after changing those 2 alignCasts! :tada:

sway ![sway](https://user-images.githubusercontent.com/10470872/177024000-23ad3e8e-6627-44f5-ba5f-6f6f291a682c.png) wayfire ![2022-07-02T23:40:40,759314746-04:00](https://user-images.githubusercontent.com/10470872/177024008-6d3150b2-3028-4055-abd6-3d95dfe534d9.png) It's running under wayland, checked with xeyes.

Requiring explicit specification of variables that can/will be used implicitly also somewhat matches things like `State` in Haskell, where you don't have to pass around a context, but it's pretty...

Maybe later, don't see a huge need for it with having landed a fix for #55 and related issues.

Yeah - the former bit is known, it's a bug in the message box library I'm using, not entirely sure how to go about fixing that for now besides rewriting...

Forgot to say this earlier, but one thing of note regarding error unions and unreachable, is that the `unreachable` case isn't actually always unreachable, annoyingly enough. If I remember correctly,...

I think *ideally*, we'd wait for #2647 - and then we could have something like ```zig pub fn read(self: *const FileProtocol, buffer: []u8) !usize { var buffer_size = buffer.len; return...