David Allsopp

Last couple of tweaking notes to those macros added. However, @shym and I discussed this separately - the benefit of the make macros is that it would catch errors just...

Assuming everything passes CI again here, and everyone remains happy, I think this is good to go, too. Relocatable and now easy to clone, too...!

This happened on #14244 as well - the AppVeyor issue is that the "CI: Full matrix" check is occasionally pushing this over its 1 hour cancellation mark on the occasions...

There's also the `--no-renames` for `git diff` which may help reduce the number of these things which appear (I think that the diff would then be a deletion of the...

Indeed, the description should be clarified. There's an explanation in the commit message which added it (see 64e08650596a67aab5654091aa721ced1cd4ea8a), but that's obviously not the best place for it: > ### Add...

Thanks for this!! That means that switches on Raspberry Pi OS now install `host-arch-arm32` instead of `host-arch-arm64` with https://github.com/ocaml/opam-repository/pull/25861. This seems a sensible change to me?

This isn't likely to come from me anytime soon - while I very much like the idea of being able to install Cygwin compilers using a native Windows opam, this...

Overlap with #29089 removed - this PR's is just waiting for me to finalise ocaml/ocaml#14250

Parking a reference to a commit which started updating this test further - https://github.com/dra27/opam/commit/f3c0b69abd2d53e8599f3c0e76548575d8aed696 (cf. https://github.com/dra27/opam/pull/22)

We could possibly look at either mechanising the relevant section or having a check on it (a la the Standard Library modules checks in the manual), given that we also...