[Releasing] rethinking the branching scheme

[Simon Hausmann]

On Tuesday 25. February 2014 12.49.23 Sergio Ahumada wrote:
> On 25.02.2014 12:00, Stephen Kelly wrote:
> > On Monday, February 24, 2014 19:26:29 Frederik Gladhorn wrote:
> >> What is problematic is:
> >> - reverse dependencies: we test qtdeclarative when changing qtbase and
> >> the
> >> other way around.
> > 
> > I still think this should be removed. It made a lot of sense before the
> > 5.0
> > release because the API in qtbase was changing so rapidly. That is not the
> > case now.
> > 
> > Unintended breaks are still caught by CI when qtdeclarative is built,
> > albeit a bit later than when qtbase is built. This is the case for most
> > repos downstream from qtbase. While it can be a good thing to prevent
> > unintended breakage earlier, I don't think it's worth it and it should be
> > removed. Use the hardware for something else instead.
> > 
> > Thanks,
> 
> I absolutely agree. revdeps should be removed.

I for one like them. They've helped us numerous times from changes landing for 
example in qtdeclarative that - had they gone in - would have blocked 
qtquickcontrols development. And we've had the same situation in qtbase - for 
build and test failures.

Pinning of qtdeclarative on the other hand due to API changes is a rare 
situation. (happened AFAICS once last year and once this year so far)


Simon