[Qgis-psc] [QGIS-Developer] Unscheduled 3.18.1 release?

[Giovanni Manghi]

Hi Nyall,



> >> Well... if you are onboard with it then I formally propose to include
> >> a 1 month testing period between the date of the creation of the
> >> installers for any .0 release and the date of the release.
> >> This would give time to test the artifacts and fix the outstanding issues.
> >
> >
>
> Two thoughts:
>
> - I'm definitely +1 to this, but I think 1 month is WAY too long. I
> would rather see a 1 week "hands off" period prior to final releases
> where we ONLY commit fixes for severe regressions which are approved
> by 2+ core developers. We did this once before, and it worked well...


It 's nice to see that there are developers and packagers that are ok
with the idea, I think we should really do it.
I proposed 1 month because my experience of testing artifacts (at
boundless) suggests that it is needed such an amount of time. My
opinion is that 1 week is not enough, but I will take it on the sport
if the other option is just nothing.





> - Isn't the Kiwi user test framework supposed to be directed at
> recognising these regressions?

yes.



>Could we get a status update on that
> and what needs to be done on pushing that effort forward? (The
> https://github.com/qcooperative/qgis-core-tests repo seems to be all
> but abandoned...?)

I'll leave to Alexandre to answer in detail, but in short the kiwi
framework would fit naturally in this temporal window between the
availability of the installers and the official release.

cheers



-- G --