Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove BUILD_DEB option #125

Merged
merged 1 commit into from
Dec 5, 2024
Merged

Remove BUILD_DEB option #125

merged 1 commit into from
Dec 5, 2024

Conversation

cajun-rat
Copy link
Collaborator

This feature allows building a Debian package that can run on a desktop machine as a kind of simulation mode. I'm not aware of anyone using this: real usages build using Yocto or run locally in development. Remove this feature.

This feature allows building a Debian package that can run on a desktop machine
as a kind of simulation mode. I'm not aware of anyone using this: real usages
build using Yocto or run locally in development. Remove this feature.
@leonheldattoradex
Copy link

We maintain a Debian package here https://github.com/torizon/aktualizr-deb (for torizon-aktualizr, but it's pretty much the same thing). Should anyone need it, it should be trivial to make it point to the upstream repo instead.

@cajun-rat
Copy link
Collaborator Author

I wasn't aware of that--if people are using it then how about bringing that upstream? The delta looks pretty minimal

@charles2910
Copy link

I think all the packaging changes are self-contained in the debian/ dir

@cajun-rat
Copy link
Collaborator Author

Are you using the Debian packing tooling rather than cpack on your branch @leonheldattoradex? If so, then I'm even more keen to nuke the stuff that is using BUILD_DEB At least then we have one maintained solution, even if it isn't upstream yet :)

@leonheldattoradex
Copy link

Are you using the Debian packing tooling rather

Yes, we have a custom pipeline that uses the usual tools[0]. One can just sbuild, debuild or else and it works, although we're moving to the more official Salsa CI GitLab pipeline [1] as soon as we have our public instance up.

[0] https://github.com/torizon/torizon-deb-ci
[1] https://salsa.debian.org/salsa-ci-team/pipeline

I'll setup a build for the "non-Torizon" Aktualizr and create an official Debian package on the upstream, what you think?

@rborn-tx
Copy link
Collaborator

rborn-tx commented Dec 4, 2024

I imagine the existing solution using cpack is something that still works, right? And it was being removed because we thought no one was using it.

@cajun-rat Now that we know someone is, shouldn't we keep it? Or does it add some relevant maintenance effort?

Replacing the cpack solution with the Debian tools seems odd to me simply because cpack is integrated with CMake which I imagine makes it easier to maintain compared with the Debian tools.

@leonheldattoradex Or are there good reasons for using the Debian tools rather than cpack?

@rborn-tx rborn-tx requested a review from jsrc27 December 4, 2024 20:12
@cajun-rat
Copy link
Collaborator Author

I imagine the existing solution using cpack is something that still works, right?

It depends what you mean by 'works' :) Running cpackgenerates a couple of .deb files, but I don't think anyone has validated that they are functional.

And it was being removed because we thought no one was using it.
@cajun-rat Now that we know someone is, shouldn't we keep it? Or does it add some relevant maintenance effort?

Replacing the cpack solution with the Debian tools seems odd to me simply because cpack is integrated with CMake which I imagine makes it easier to maintain compared with the Debian tools.

@leonheldattoradex Or are there good reasons for using the Debian tools rather than cpack?

The current situation as I see it is we have 2 independent implementations of similar functionality:

  • Upstream uses cpack, but no-one uses it
  • @leonheldattoradex maintains a downstream version that uses the Debian tools

Since @leonheldattoradex is the active maintainer here, I'll let them pick the implementation approach. I recall the Debian community prefers to use Debian tooling to build packages rather than those built into other tools, so while cpack is good for people who are doing 'cmake' it is worse for people doing 'Debian'.

I think there are two approaches from where we are today:

  • Nuke BUILD_DEB now, and incrementally bring torizon-deb-ci closer to upstream.
  • Migrate torizon-deb-ci to cpack and upstream it now instead of this PR.

My preference would be for the first, but if someone wants to work on the latter I'm happy to help!

@rborn-tx
Copy link
Collaborator

rborn-tx commented Dec 5, 2024

It depends what you mean by 'works' :) Running cpackgenerates a couple of .deb files, but I don't think anyone has validated that they are functional.
[...]
I recall the Debian community prefers to use Debian tooling to build packages rather than those built into other tools, so while cpack is good for people who are doing 'cmake' it is worse for people doing 'Debian'.

Ok, considering the above, it seems reasonable to nuke BUILD_DEB. Approving the PR.

@cajun-rat
Copy link
Collaborator Author

Cool. I'll merge this now. Shout if I can help with the upstreamng bit!

@cajun-rat cajun-rat merged commit 6e90a8d into master Dec 5, 2024
5 checks passed
@cajun-rat cajun-rat deleted the tidy/no-build-deb branch December 5, 2024 18:55
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants