Gabriel Scherer
Gabriel Scherer
You removed the commits I was grumpy about (thanks!). I had a new look and they are overall fine, but I don't know about the last one ("extract common code...
> It's also just possible to add prototypes for `bt_lock_{acquire,release}` and `st_masterlock_waiters` in st_pthreads.h and put the functions definitions in st_stubs.c. Would that be ok with you? I find this...
> I don't have a better idea on how to improve on that. Have you considered the suggestion I made [earlier](https://github.com/ocaml/ocaml/pull/13352#issuecomment-2299316030)? It was as follows: > Could the prototype `void...
You forgot to update the Changes entry to credit @dustanddreams' bug-catching review.
Merged, thanks!
I broadly agree with @yallop. I'm not fond of adding more subtyping relations for low-level efficiency hacks, but I would be in favor of ensuring that conversion functions written as...
> Is ocamldoc still actively used and maintained, or is everyone using odoc (which doesn't have this issue)? I can look into fixing it, but it will require rewriting most...
cc @Julow this is non-urgent, but I wonder if you would be interested in reviewing this
> Lidents could possibly be passed to the polymorphic compare, which can now return a different result. I couldn't find a Map or a Set on long idents but I've...
I think that the amount of testing on this question (changes in polymorphic comparison behavior) is now adequate, we can consider that the current behavior is correct. This needs a...