| 1 |
On 9.3.2022 23.00, Matt Turner wrote: |
| 2 |
> I'd like to deprecate and ultimately remove repoman. I believe that |
| 3 |
> dev-util/pkgcheck and pkgcommit (from app-portage/mgorny-dev-scripts) |
| 4 |
> are far superior replacements, and it makes sense to have people using |
| 5 |
> the same tool and seeing the same warnings as in the CI. |
| 6 |
> |
| 7 |
> Are there any useful checks or behaviors of repoman that are missing |
| 8 |
> from pkgcheck and pkgcommit? |
| 9 |
> |
| 10 |
> Thanks, |
| 11 |
> Matt |
| 12 |
> |
| 13 |
|
| 14 |
I still fail to see the "why" in here. Repoman is better than pure 'git |
| 15 |
commit' that I know some people still like to use, and as long as it's |
| 16 |
kept maintained. |
| 17 |
We should be teaching people about the alternatives, and let them choose |
| 18 |
whatever they like more. |
| 19 |
https://wiki.gentoo.org/wiki/Package_maintainer%27s_responsibilities#Ebuilds_and_git_workflow |
| 20 |
|
| 21 |
Repoman is a very lightweight tool. All that being said, I can't think |
| 22 |
of a single feature that pkgdev+pkgcheck don't cover when switching from |
| 23 |
repoman. |
| 24 |
|
| 25 |
The quick global CI checks after each commit saves us from a lot of |
| 26 |
trouble. If you do bad commits, you get immediately noticed about it and |
| 27 |
you can fix it rather quickly usually. When you "get hit", you also |
| 28 |
learn about pkgcheck and we've seen that this is the point when people |
| 29 |
usually integrate it to their workflows. I'd also like to point out |
| 30 |
whenever doing more complicated pushs, one can trigger the automatic CI |
| 31 |
check in our Github mirror via a pull request before pushing. |
| 32 |
|
| 33 |
My only worry is: are pkgcheck and pkgdev _really_ maintained anymore? |
| 34 |
More than repoman is? |
| 35 |
|
| 36 |
-- juippis |
Archives