Packit build failures after merge

[praiskup]

We have "negative" CI karma for quite some time:
https://github.com/fedora-copr/copr/commits/main

What if we disabled Packit for the time being for pushes?

cc @lachmanfrantisek @lbarcziova

[lbarcziova]

hi @praiskup ! Are all the failures false positives?

[nikromen]

I can see failed builds since fc73a83

perhaps it's due to the ~~dev change we did in copr-common version

edit: I guess grep '"%version"' setup.py fails when build is submitted by packit

[nikromen]

So it seems like parsed spec file in https://download.copr.fedorainfracloud.org/results/packit/copr/fedora-39-x86_64/06266070-python-copr-common/ and its SRPM is different (look to the spec file in SRPM and spec file inside .tar.gz in SRPM - their version differs) - that's responsible for the majority of errors

another bug I found (some time ago) and is responsible for the first fail in https://copr.fedorainfracloud.org/coprs/packit/copr/package/python-copr-common/ is that versions in spec file in SRPM and in tar.gz (inside SRPM) sometimes differ - usually when I submit a PR with new pkg version then it has still its lower version and one more push to PR/new PR to catch up/refresh the version is needed

but I have no idea whether this is our or packit/specfile issue

[nikromen]

also, I don't think it's a ci-blocker since we have our CI still running in parallel

[praiskup]

edit: I guess grep '"%version"' setup.py fails when build is submitted by packit

For some reason, Packit generates version 0.20.1, while our specfile has 0.20.

also, I don't think it's a ci-blocker since we have our CI still running in parallel

Well I need to have green CI mark in recent commits, otherwise it blocks my brain. :-)

[praiskup]

E.g. here it seems that wait-for-copr doesn't actually wait? Because copr-backend is built before python-copr-common.

[praiskup]

#2956 likely fixes this!

[praiskup]

And all green, finally!

[praiskup]

Thanks @lachmanfrantisek :)