https://bugs.kde.org/show_bug.cgi?id=454345
Uwe Dippel <udip...@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |VERIFIED --- Comment #69 from Uwe Dippel <udip...@gmail.com> --- (In reply to Nate Graham from comment #66) > Because we know this issue has multiple underlying root causes, we need to > treat each one individually. We think it's fixed in Plasma 5.27.3 at this > point. William hasn't told us otherwise; his last comment concerned Plasma > 5.26.4 which we know was still buggy. > > If you're saying that the issue still happens for you when you run > `kquitapp5 plasmashell && kstart5 plasmashell`, we need to investigate that > separately, because it could be (and probably is) happening for a totally > different reason from why it was happening for William. It's totally > possible that William will say, "hallelujah, it finally works for me in > Plasma 5.27.3" and you'll say, "grr, no it doesn't, running `kquitapp5 > plasmashell && kstart5 plasmashell` still breaks it for me!" All that means > is that the same proximate causes (restarting plasmashell using kquitapp5 > and kstart5) trigger different codepaths and you're hitting a buggy one. > Hence, we need to fix that separately and we need a new bug report for it. > > --- > > But if you want to know what will *actually* happen next, it's that > developers will largely ignore the issue because they're completely > overloaded with work and this issue is non-deterministic and impossible to > reproduce at will. Eventually developers woh do get some time to work on > this will throw out the old code because at this point we've learned that > it's fundamentally flawed, and they write a new system that is > architecturally better, like we recently did for KScreen. In the process > they will fix the root causes of this issue but introduce new ones that were > unanticipated or are the result of imperfect config migration, so the system > will still feel buggy especially to existing users, just in different ways. > Eventually those issues will be fixed one-at-a-time as well. After 2 or 3 > years of bugfixing a system with a sane and comprehensible architecture, > this part of the system will finally feel stable to new users and old users > who have re-installed or figured out how to reset their configs. > > Anything other than the above will require more development and QA > resourcing than currently exist. If you would like to help make this better, > I would encourage contributions with code, QA, bug triaging, or donations. > Anything else isn't helpful, sorry. If you want to help, then help. Don't > just tell us we're doing it wrong. What we need is more contributions from > people who can help fix things. Yep. My coding isn't current, alas. I have offered help multiple times, though not on the level of code. -- You are receiving this mail because: You are watching all bug changes.