ocaml 5.2.0

[XVilka]

• Have you followed the guidelines for contributing?
• Have you ensured that your commits follow the commit style guide?
• Have you checked that there aren't other open pull requests for the same formula update/change?
• Have you built your formula locally with HOMEBREW_NO_INSTALL_FROM_API=1 brew install --build-from-source <formula>, where <formula> is the name of the formula you're submitting?
• Is your test running fine brew test <formula>, where <formula> is the name of the formula you're submitting?
• Does your build pass brew audit --strict <formula> (after doing HOMEBREW_NO_INSTALL_FROM_API=1 brew install --build-from-source <formula>)? If this is a new formula, does it pass brew audit --new <formula>?

---

[github-actions]

Thanks for contributing to Homebrew! 🎉 It looks like you're having trouble with a CI failure. See our contribution guide for help. You may be most interested in the section on dealing with CI failures. You can find the CI logs in the Checks tab of your pull request.

[chetmurthy]

I'm not sure where the test is in the brew build process, but the problem is that where you load "pr_o.cmo", now you must load "o_keywords.cmo pr_o.cmo". But the old way of invoking camlp5 for preprocessing (e.g. "camlp5 pa_o.cmo pr_o.cmo") is cumbersome and predates findlib. Instead you can use "not-ocamlfind preprocess -syntax camlp5o -package camlp5.pa_o,camlp5.pr_o") as is done here: https://github.com/drjdn/p5scm/pull/4/files

[chetmurthy]

Yes, "pr_op.cmo" provides extensions to the "pr_o.cmo" printer. The entire methodology of loading individual .cmo files into the camlp5 process is flawed, b/c you have to specify all the files you want, and in a compatible load-order. There is no way to get inference of which other files are needed, for the files you specify, etc. All of this sort of thing is the proper domain of findlib, which knows exactly these things and does it when constructing preprocessor commands -- so that is what not-ocamlfind preprocess hooks into.

[XVilka]

Yes, "pr_op.cmo" provides extensions to the "pr_o.cmo" printer. The entire methodology of loading individual .cmo files into the camlp5 process is flawed, b/c you have to specify all the files you want, and in a compatible load-order. There is no way to get inference of which other files are needed, for the files you specify, etc. All of this sort of thing is the proper domain of findlib, which knows exactly these things and does it when constructing preprocessor commands -- so that is what not-ocamlfind preprocess hooks into.

Thanks for pointing this. This is my first time updating Homebrew package, so I don't know why this particular test was written this way, but from the comment purpose it was made to check the binary compatibility with the OCaml compiler. I will try to change to that not-ocamlfind instead.

[chetmurthy]

Ah, you're trying to check that the output of camlp5 is binary-compatible with the ocaml compiler ? In that case, you shouldn't be loading "pr_o.cmo" -- that produces text source. If you want to check binary compatibility, you could use not-ocamlfind preprocess -syntax camlp5o -package camlp5.pa_o,camlp5.pr_dump) . This would output the result of preprocessing as an OCaml binary AST.

[chetmurthy]

If the issue here is tests, you could use the extensive test-suite that comes with Camlp5. When I took over maintenance, the test-suite was pretty limited, but I've added a massive set of tests of all sorts of function. You can look at the opam file to see how they're invoked. And if you have other tests in mind, I can add those too!

[XVilka]

Ah, you're trying to check that the output of camlp5 is binary-compatible with the ocaml compiler ? In that case, you shouldn't be loading "pr_o.cmo" -- that produces text source. If you want to check binary compatibility, you could use not-ocamlfind preprocess -syntax camlp5o -package camlp5.pa_o,camlp5.pr_dump) . This would output the result of preprocessing as an OCaml binary AST.

For now I use the old solution, since the not-ocamlfind is not available at this stage.

[carlocab]

Thanks!

[XVilka]

Strange error while building Coq:

  ==> ./configure -prefix /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1 -mandir /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1/share/man -libdir /home/linuxbrew/.linuxbrew/lib/ocaml/coq -docdir /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1/share/coq/latex
  ocamlfind: Config file not found - neither /home/linuxbrew/.linuxbrew/etc/findlib.conf nor the directory /home/linuxbrew/.linuxbrew/etc/findlib.conf.d
  Error while running '/home/linuxbrew/.linuxbrew/opt/ocaml-findlib/bin/ocamlfind query findlib -format %v' (exit code 2)
  Configuration script failed!

[XVilka]

Strange error while building Coq:

  ==> ./configure -prefix /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1 -mandir /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1/share/man -libdir /home/linuxbrew/.linuxbrew/lib/ocaml/coq -docdir /home/linuxbrew/.linuxbrew/Cellar/coq/8.19.1_1/share/coq/latex
  ocamlfind: Config file not found - neither /home/linuxbrew/.linuxbrew/etc/findlib.conf nor the directory /home/linuxbrew/.linuxbrew/etc/findlib.conf.d
  Error while running '/home/linuxbrew/.linuxbrew/opt/ocaml-findlib/bin/ocamlfind query findlib -format %v' (exit code 2)
  Configuration script failed!

This looks similar to Homebrew/homebrew-test-bot#805 4f236f2, just different paths.

[carlocab]

Test failures other than hyperkit look unrelated.

[chenrui333]

@XVilka, thanks for your contribution to Homebrew! 🎉 🥇

Without awesome contributors like you, it would be impossible to maintain Homebrew to the high level of quality users have come to expect. Thank you!!!!

[github-actions]

@chenrui333 has requested bottles to be published to this PR.

[chenrui333]

need to followup on this

[chenrui333]

https://github.com/Homebrew/homebrew-core/actions/runs/9395167391

[XVilka]

Hyperkit really looks like abandonware now:

• https://github.com/moby/hyperkit/commits/master/
• https://github.com/moby/hyperkit/issues
• https://github.com/moby/hyperkit/pulls

Not much could probably be done for it, until developers/maintainers step in.

[XVilka]

Apologize for tagging you @djs55, but since you were hyperkit developer, you might be able to help reviving it.

[chenrui333]

yeah, there are no new commits for the past two years.

[avsm]

It's not so much abandonware, but stable. There's no need to keep it up-to-date with the very latest OCaml, and I doubt there are many users accessing it via Homebrew. I'm happy to it to be removed from Homebrew to ensure there's no blocker on upstream OCaml updates. Copying in @djs55 as another maintainer of Hyperkit to confirm.

[djs55]

I would describe it as stable -- it's still shipped in Docker Desktop for Intel Macs. Removing hyperkit from Homebrew is fine.

[github-actions]

❌ @chenrui333 bottle request for hyperkit failed.