truemedian

I think the actual issue here is that fetching a private repo requires a bearer token with an additional scope. Currently, gyro's bearer token is only used for reading user...

I have written a library that does just this. It is [truemedian/zfetch](https://github.com/truemedian/zfetch). It only requires: - [MasterQ32/zig-network](https://github.com/MasterQ32/zig-network), to provide a system-agnostic socket layer - [MasterQ32/zig-uri](https://github.com/MasterQ32/zig-uri), for uri parsing - [truemedian/hzzp](https://github.com/truemedian/hzzp)...

> I assume this is being worked around via disable_sanitize_c = true? It *should* be, but yes, its still emitting `ub1` instructions and I didn't take the time to figure...

I figured out why luajit was being ubsan'd (the library itself needed no-sanitize-c), however, something in libc's atexit is causing a segfault, and I can't figure out why that would...

Whoops, looks like I forgot the newline at the end of the file :sweat:

The issue with running tests on them is that there is no testing infrastructure for uefi, there is simply no way to run them without doing something weird with the...

It would appear that the change making packed structs backed by integers also made them no longer align(1). So I guess I also need to go through and align(1) all...

I've cleaned up the commit history, if people can verify that it also works for them it seems like this can be merged.

I finally found the important words for DevicePathProtocol. > Therefore all nodes are byte-packed data structures that may appear on any byte boundary. All code references to device path notes...

I think everything that broke has been fixed, lgtm