New packages: mpm-7.6.1, python3-click-extra and python3-extra-platforms

[kdeldycke]

Summary

This PR adds Meta Package Manager, a Python CLI which wraps multiple package managers (apk, apt, brew, npm, pip, xbps, ...) behind a unifying interface. It already wraps XBPS itself in meta_package_manager/managers/xbps.py.

Testing the changes

• I tested the changes in this PR: YES

Built each on a Void x86_64 VM against master with ./xbps-src -Q pkg.

xlint reports nothing for any of the four templates.

I used a public CI workflow to build the entire dependency tree end-to-end so you can inspect the results: https://github.com/kdeldycke/meta-package-manager/actions/runs/25785905046/job/75739147510

New package

• This new package conforms to the package requirements: YES

This introduce 6 new pure Python packages:

• 3 I do not own and I proposed separately in: New packages: python3-{cloup,deepmerge,packageurl-python} & python3-boltons update #62477
• 3 I am the author of:
  • meta-package-manager itself
  • python3-click-extra dependency
  • python3-extra-platforms dependency

Local build testing

• I built this PR locally for my native architecture, (x86_64-glibc)
• I did not build it for other architectures: all packages are pure Python and ship no compiled code.

Context

I am submitting this package because mpm has external users on other distributions (Homebrew, AUR, stew, with pending submissions to Nixpkgs and GNU Guix) and there is no community packager currently working on Void. All three packages have been published to PyPI for years and have third-party consumers.

Also, the reason the Void Linux project landed on my radar is because a user send me a PR on my extra-platforms project to add that distribution to my inventory of detections (see: kdeldycke/extra-platforms#523 by @stiermid). So I guess it is a good opportunity to integrate better with your project.

[kdeldycke]

I just simplified the dependencies thanks to the 6.6.0 new release. Is is poosible for a maintainer to allow the workflow to run so I can hunt for packaging issues?

[nazaradeh]

Sadly as you mentioned in your caveat, Void does not allow the developer of the software to package their own software here.

It's also encouraged that someone submitting a new package to have contributed in the past (e.g. updating orphaned packages):

To ensure packages stay maintained, we prefer that they are submitted by those with a history of contributions to existing packages. A good way to do this is to work on orphaned packages (those maintained by orphan@voidlinux.org) that you use.

If you have someone who uses mpm on Void, encourage them to first contribute to the repo with package updates, then open a PR to submit mpm as a new package themselves. If you can't find anyone, I can consider opening a PR myself, I just like to limit my maintained packages to things I use myself.

[nazaradeh]

• Add python3-packaging to depends

PYTHON: py3:packaging <-> python3-packaging NOT IN depends PLEASE FIX!
=> python3-wheel-0.47.0_1: missing dependencies detected!
=> python3-wheel-0.47.0_1: please add these packages to depends: python3-packaging

• Any reason for your post_install() to be the way that it is? You just have a license file, so I figure your post install should just be this:

post_install() {
	vlicense license
}

• The above is not relevant to this next point, but this package fails to build on my system (x86_64)

/usr/bin/ld: /usr/lib/libarchive.a(archive_cryptor.o): undefined reference to symbol 'EVP_EncryptUpdate@@OPENSSL_3.0.0'
/usr/bin/ld: /lib64/libcrypto.so.3: error adding symbols: DSO missing from command line
collect2: error: ld returned 1 exit status
ninja: build stopped: subcommand failed.
=> ERROR: cmake-4.2.2_2: do_build: ... exited with 1

I don't have time to look further since it built for an hour before it reached this error. Were you able to build it on your own Void machine?

[kdeldycke]

Sadly as you mentioned in your caveat, Void does not allow the developer of the software to package their own software here.

I completely understand the decision. And I've read the contribution guidelines before proposing this PR so that's why I was upfront and pointing to that violation of your rules.

In the meantime, and to salvage the time I already invested in trying to contribute to the void project, I proposed another PR at #62477 . This one contains the update and new packages I am not the author of. So you can evaluate them independently. As a show of good will, I proposed to take over the maintenance of the orphaned python3-boltons package.

If #62477 is merged, I will rebase this PR on top of master so it will only contain my proposal for my own software.

Now, is there a way to satisfy the bureaucratic process? @stiermid are you interested to propose the extra-platforms package that you use? I authorize you to steal from this PR and change my name to yours.

Another way to solve this: maybe @nazaradeh you can just duplicate this PR, propose it under your name and once it is merged into master I can take over its maintainership?

[kdeldycke]

• Add python3-packaging to depends

You're right. And this has already been fixed in master by c24683f . So rebasing on it fixed the issue.

• Any reason for your post_install() to be the way that it is? You just have a license file, so I figure your post install should just be this:

This is just me imitating some other Python packages and was under the impression it was somewhat mandatory. Just running xlint I realized I do not need them. Fixed.

• The above is not relevant to this next point, but this package fails to build on my system (x86_64)

I don't have time to look further since it built for an hour before it reached this error. Were you able to build it on your own Void machine?

I just rebased my PR on master, updated the dependencies and lint errors, and it fix the issue.

[kdeldycke]

And thanks @nazaradeh for the feedback! :)

[ahesford]

Our general objection to new-package submissions originating from an upstream author is not a matter of bureaucratic process. It is an effort to ensure quality and deny a venue for self promotion. I do not appreciate your attempt to subvert our standards.

If regular Void contributors find your package useful, let them propose its addition. This also provides a means for third-party validation of the suitability of your package and its further updates for distribution to Void users.

@nazaradeh please refrain from advocating that contributors subvert our expectations for submissions or offer your assistance in doing so.

[kdeldycke]

@ahesford I apologize. It was not my intention to subvert your process. I am myself a core contributor to a popular package (that you are probably using everyday) and understand your position. Burn out is general in the OSS world because of low-quality contributions. My previous comments are more like a call for external contributors to take over my work. Contributing to exotic distributions is a good way for me to debug and improve my software. And sorry if you had an impression of subversion: I might have used harsh words unintentionally as English is not my mother tongue, so some subtleties might have escaped me. All in all, I accept your decision to close this PR.