David Allsopp

I think one's good to go as steps in the correct direction? (and eliminating that pesky warning!)

Your review & approval on these things is much appreciated, thank you!

This can indeed be merged now - but can the `setenv` and `x-env-path-rewrite` also be removed, as per #29085?

Yup - it remains necessary for ocaml-system, where it's then being used "correctly" (because we actually can't alter the compiler)

It looks like you've generated `configure` by manually running `autoconf` - could you rebase the commits but use `tools/autogen` instead, please? (I'm happy to do it, if you prefer). This...

Sorry for the delay - can't be absolutely certain, since the full build log isn't there, but I'm 99% certain that this is https://github.com/ocaml-windows/papercuts/issues/2. The other workaround is to do...

It's currently the case and I'm fairly sure it was the original intention (@AltGr?), that `ocaml-system` is related to `ocaml-base-compiler` - i.e. that you get a _vanilla_ OCaml compiler. That's...

The difference with both of the approaches I'm suggesting is that it only works arbitrarily for branches (with the pinning approach) and only _during_ the release cycle for the pre-releases...

Perhaps to be less unclear, what I’m concretely suggesting is that there’d now be no change to ocaml-system.5.4.0 here (because it’s released), but to add ocaml-system.5.5.0 largely as here (just...

Does that fit in with [this related thought](https://www.dra27.uk/blog/platform/2025/09/30/file-descriptors-are-not-integers.html)? Would it be the case that you _must_ pass the fd to a program with its current number, or could you pick...