@pony thanks, I'm glad I'm not the only one suffering
so you don't, all that minor shit adds up and then you need to do an update for a new feature or something and it just goes to hell and everyone's getting sad, why haven't we been keeping it up to date, thus the cycle continues
@pony regressions are evil
@wolf480pl we have Jenkins, Travis, GitHub Actions and CircleCI. None of them are great, and TBH none of them seem clearly superior to the others either...
I wouldn't say dumpster fire, we do fix things up to the point where it's not *too* painful... though never to the point where it is pleasant ;).
@raboof so you found a way to make it not accumulate 10s if gigabytes of garbage, hacks, and obsolete stuff over the years?
@wolf480pl tbh no I didn't, I just stopped caring too much :D.
I'm still undecided whether I find the "fix just the things that hurt most" approach actually good or not: it kinda works but also it means the problems that don't hurt enough don't get fixed.
I've found having a hardware failure is a good way to clean some things up :D
@wolf480pl at KDE we are soon getting rid of that dumpster fire.
@wolf480pl Mixxx used to use Jenkins. It was awful. Keeping the servers running was a huge distraction, the people running them wouldn't allow anyone else SSH access to fix the regular problems, and they'd disappear for weeks or months at a time. Using GitHub Actions is way easier and more reliable.
@[email protected] I'm looking for a new job, one company bragged that they have a >1000 large jenkins cluster. 🚩🚩🚩
