[QGIS-Developer] A plea: more volunteers needed for reviewing backports

[Alexis R.L.]

Greetings,

I'm curious to know if all backport should be reviewed. In theory other
than major code rework backports should be retrocompatible. Assuming they
are most often done by domain experts as is mostly the case, would it be
too risky to merge if the main PR was reviewed and the tests are all green?

I'd agree there should be exceptions such as after major refactoring (if
exclusive to some versions).

Or could there be some way to simply identify backport that would require
review and those that would not affect more than the intended area? I'm
also curious to know if there could be a way to rely less on the few
highly-skilled contributors that are active when it comes to more minor
things (if there are such a thing).

Thanks as usual and have a great day,

Alex


Le lun. 22 mars 2021 à 19:07, Nyall Dawson <nyall.dawson at gmail.com> a
écrit :

> Hi list,
>
> This is a public plea for more developers who are very familiar with
> different parts of the QGIS codebase to become actively involved in
> backport PR management.
>
> We NEED more people to be actively (i.e. daily) monitoring for
> backport PRs and then reviewing them if the backport affects code
> areas which they're knowledgeable about.
>
> Right now there's like... 2... of us doing this on a daily basis, and
> we need more hands and eyes. Regressions are fixed by backports, but
> they can also be introduced by backports. If we don't get more people
> involved in this part of QGIS' maintenance then the stability of the
> patch releases is compromised.
>
> While I'm more than happy to keep reviewing backports which touch
> areas of code I know well, we have a whole stack of backports for
> server fixes, relation fixes, and other parts of QGIS I don't know
> well enough to risk release stability by reviewing myself. These
> backports are currently sitting in the queue for weeks without review,
> and often they stale out and are closed without ever getting merged.
>
> I've also a ton of backports for fixes that I make on a daily basis
> which I cannot self review, so the PR queue gets flooded with my
> backports until I actively pester/harass others to review and approve.
>
> The truth is that the longer the PR queue, the more it stresses me out
> and the less motivation I have to review the incoming feature PRs. We
> have many older feature PRs which are stalled and waiting review for
> months or longer, just because I'm so flooded with managing newer
> incoming PRs that I never have time to review the older ones...
>
> So please, if you know parts of the QGIS code well, and can spare 15
> minutes a day to looking over the backport queue and approving any
> backports you are confident are regression free, then please get
> involved! (or ask your employer to officially allocate you this 15
> minutes/day as a way of them supporting QGIS!)
>
> Nyall
> _______________________________________________
> QGIS-Developer mailing list
> QGIS-Developer at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-developer/attachments/20210323/37b71397/attachment-0001.html>