Hello people, I said it previously but I'll repeat it again: --- If build.kde.org is not green for kdelibs-frameworks or plasma-framework build, no changes should get in apart from attempts at fixing the problem. ---
So far, it has been completely ignored and that's really a bad situation. We made lists of tasks, the goal is not to take as many as possible in parallel and throw patches at them as fast as possible. The goal is to get through them while maintaining or increasing the quality. Our build history tells me there's either a general lack of disdain for quality, or some of us don't know how to look at the clues we have. I've been spending most of my time on KF5 lately just fixing builds. It's not good and not warranted. This week is really over the top in breakages: * yesterday we had more than a hundred of our test suites not running on a make test (I'm still not done having all of them back, still investigating); * today I still have test regressions in two test suites (as can be seen on build.kde.org, still investigating). The issue is not who or what introduced those problems (some of them likely come from the switch to qt5/stable), the issue is that we are ignoring build.kde.org reports. That has to change, now! So here is a list of ideas on how to improve: * if you are merging a branch or moving something around: then make a clean build (removing both build and install dirs) before your push, you will likely catch something; * if you see that build.kde.org is not green: then try to find a solution, don't just keep pushing patches which will introduce further instabilities; * if you see that build.kde.org is not green: then don't post or review patches which are not working toward getting the build in shape again. In case you didn't notice, I've been applying all of that myself (for instance I reviewed exactly one patch today because it fixed a regression, ignored everything else). I'm making myself slower but it's the speed of the team which matters, and rework just slows the whole team... Obviously, for the time being I don't want to see any push in kdelibs frameworks which doesn't aim at fixing one of the problems we have in the build currently. Thanks for your attention, I hope I won't need to bring up this topic again. Regards. -- Kévin Ottens, http://ervin.ipsquad.net Sponsored by KDAB to work on KDE Frameworks KDAB - proud supporter of KDE, http://www.kdab.com
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ Kde-frameworks-devel mailing list [email protected] https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
