Today work has focused on continuing to review GitHub notifications on rust-lang/libc, and on working through some more of the SML for the working programmer book.
I started off by looking through a bunch of PRs and issues I was mentioned in. I’m still not done with all of them, but the list has gone down to 4 pending notifications.
Among those I did get to address today, I’ve finished revising the PR that implemented support for the exhaustive attribute, commented with feedback of my own on another contributor’s PR, and rebased the uClibc PR for LFS support.
The latter I’ve done twice since last week even though there was no explicit feedback. This is because the one target maintainer that did comment on the PR thread mentioned they’d be looking into it this week, and there were conflicts.
The PR I provided feedback on was one concerned with removing sighandler_t as an alias on platforms that do not use glibc as their environment. I commented on two things; Moving stuff over to the new module and looking at Windows.
The patchset moved around a bunch of definitions and FFI symbol bindings, which seemed to me like we could altogether move into new. Then I remembered that there is one target using glibc that was not considered; MinGW.
In terms of compiler work, today I’ve only continued working through the SML book I picked up yesterday. I am increasingly doubting whether I should continue working with any ML language, as I have recently also looked into Idris.
I have also spent some time looking into another browser with a fairly strong anti-AI stance, but had no chance of finding one. The closest thing is Waterfox, which is sort of there; I’ve switched to it because Safari is too painful.
None.
The rust-lang/libc work on the ctest extension will resume once I have finished looking through the PRs I got feedback on. Only the general time64 cfg patchset and the MCP implementation remain; I should be done tomorrow.
I was also pinged on a tracking issue for platforms where we may still want to make changes concerning time64, just to check whether I think something is missing. That I’ve yet to go through.
Another PR I’ve yet to open as follow-up for completing the siginfo_t type definition across targets is to change the skips in the build script. This is still pending.
Compiler work will likely be put on halt temporarily. I want to explore the Idris documentation, and see into both using it for the compiler and properly learning a proof assistant. It is pure and theorem-oriented, so that’s a plus.