This would be why I moved kdiff3 to Qt6 only in the latest release. I took the ci breakage on Qt5 craft as a warming that the time was coming. Additionally I no longer have a proper way to test against Qt5 beyond build-ability.
Dec 10, 2024 11:38:38 AM Volker Krause <vkra...@kde.org>: > On Sonntag, 8. Dezember 2024 11:56:19 Mitteleuropäische Normalzeit Albert > Astals Cid wrote: >> El dissabte, 7 de desembre del 2024, a les 14:01:58 (Hora estàndard del >> Centre >> d’Europa), Robby Stephenson va escriure: >>>> Probably worth reviewing in more detail where those are on the way to >>>> Qt6 >>>> and >>>> what can be done to support/accelerate the transition. >>> >>> In Tellico's case, it's fully ported to Qt6. It still compiles on Qt5, >>> though, which is why the CD job hasn't been removed. I figured I'd keep it >>> running until it wasn't supported. >> >> I would suggest you only support one of the Qt versions and if Qt6 is >> supported I would suggest removing the Qt5 one, let's not give us more >> support problems than the ones we need. > > Agreed. Dropping Qt5 support in places where a working Qt6 port already exists > and the Qt5 support isn't strictly necessary (like it is for the style or > platform integration plugins) would seem like an easy and useful step forward. > That would make it easier see the real blockers and would reduce the > maintenance and infrastructure load. > > Besides the two mentioned CD users, there seem to be a few more cases like > that among the CI users as well. The attached list contains all repos with Qt5 > CI/CD jobs that also have at least one Qt6 job. > > Regards, > Volker
