Rebase const-generics branch
Closed this issue · 4 comments
@yodaldevoid: I'd like to rebase the const-generics branch on to master to fix up the few merge conflicts that have arisen, and also avoid having to rebuild the new version of LLVM after switching between const-generics and other branches each time. I was originally going to wait until rust-lang#48149 had been merged, but as that'll probably take a little while longer, I'd rather do it first now.
Can I force push a rebased version of the branch, or will that cause issues for you at the moment?
Feel free to go ahead. I probably won't be able to work on this for a few weeks as I am currently swamped by school work.
I've been holding off on making more changes to const-generics until the two other refactoring PRs get merged. After that, though, there are only a few TODOs left — though I ran into some compilation errors when trying to compile the examples, so there are definitely still some rough patches that need to be looked into before we have something that works.
@yodaldevoid: just thought it'd be good to update you on where I am with this (though you've probably noticed the lack of movement) — I've been busy and I'm not going to have time for the next month and a half to work on const generics, other than continuing to push the refactoring PRs through. In about 6 weeks, though, I should have a lot more time and plan to start working hard!
Just from some of the rebasing I did after miri: there's quite a lot that's changed, both for better (some of our // TODOs are irrelevant now) and for the worse (fixing everything) — and this will be even more true after the two generics PRs get merged, so it might even be more convenient overall this way.
I'm not sure when the lazy normalisation problems will hit (it seems there are at least some uses of const generics that should be fine), but integration with Chalk is already beginning to start, so it might even line up well with our branch. I'm optimistic we could be starting to merge things in by the end of summer...
I've been keeping an eye on the refactoring PRs. I should finally have time to work on this for long enough to move it forward around the same time that the first one goes through.