Gräßlin: Multi-screen woes in Plasma 5.7

I filled some bugs with as much info as I could, and these collected lots of me-too responses, more info and suggestions. The bugs seemed to snowball, but there was never any progress or dev involvement so the energy was wasted.

I moved to LXDE, but then with HiDPI the whole thing got just too bad again. A new laptop later and I actually use Win10 with Linux in VirtualBox. It works well enough, and I can get Vbox to do screen scaling too if needed.

The sad thing is that I remember multi screen working really well in KDE on fedora 13 or something around then. I remember undocking, all the windows coming onto the panel, then docking again and them going back to their original multi-screen layout. Hopefully we will get back to that sort of thing one day again.

1) Why was it released to users without testing? It always take me 20 sec on any KDE fresh install to make it crash very badly, with multi-screen.

2) Ok, the underlying stack is rotten. But why is it working with Gnome? If you haven't yet, you should test this environment and pick some idea. At least they respected their users in finding a workaround.

So, yes, pointing out the weaknesses is good, but coming with a solution is better.

Because, even though I like Plasma, it is useless to me as is, so I am using Gnome.

1. ignored comment, did not publish it, because it pointed out a series of core problems with your approach to software development and bug reporting, using specific examples of massive bugs and failures from the 5x release to highlight the points, I assume that's why anyway.
2. locked thread for further comments
3. ignored another comment I put on another posting, same story.

My conclusion was this: kde is broken internally as a project and is currently unable to ship quality software that can be trusted. I moved to xfce, which thank god still appears to be run by responsible competent adults. If you're confused about how to properly integrate a new toolkit, look how xfce is doing it with gtk3. Slowly, bit by bit, probably waiting for issues to be resolved in gtk3 (and believe me, I'm NOT a gtk fan), then rolling out support for the new toolkit, once it's stable. Not breaking it all then making excuses, which seems to be the too big and pointlessly bloated to succeed model that gnome and kde are both following.

I further concluded that the kde project itself is the problem at this point, not a specific issue or bug, certainly I couldn't trust kde to deliver a desktop I'd want to use, even when working, it was like you guys never even try to stress your desktops, so the stuff is incredibly slow, particularly when you load it with lots of programs and virtual desktops (something it appears you guys don't even use as developers anymore given how incredibly poorly these things ran in 5.x and 4.x to a lesser but still quite poor degree), I had accepted this slowness in kde 4.x, but when I went to xfce, I suddenly saw how bad kde had gotten, switching virtual desktops now works just the same as it used on kde 3.5, when it was fast, like a desktop should be.

It seems like you guys have gotten lost in all your glitzy details, and actually believe that a desktop is something more than what runs your applications in a smooth, fast, clean way. it's sad, because a theoretically good and working kde would still be the best out there, but kde hasn't been good and fast for years, and 5.x isn't even working, though there's lots of excuses given.

I don't blame you for being defensive, that's certainly the easier option, rather than actually looking at the kde project and trying to figure out how things can have gotten so distorted that basic functionality is left to rot and die while you do all kinds of exciting things like adding support for wayland, something that barely even exists out here in the real world yet.

My conclusion is that there was never enough developer mindshare to maintain two distinct full featured desktops, and there still isn't, and there certainly isn't enough to maintain massively bloated stuff like kde 4-5x have become, or gnome, or whatever. Incremental development would have been the key, with well planned gradual integration of new toolkits, but that would probably require a mindset your project no longer has access too, which is too bad.

And I speak as someone who spent years doing heavy duty active support for a kde based distro, and really wanted to like kde but in the end I had to wave the white flag and give up, it simply was not practical to deal with software from a group that simply seems unable to do anything remotely close to good programming. kids, lol, what can you say, you can only get what you get out of college.

I'd tentatively suggest that a few simple tests be used to check design concepts:

1. the virtual desktops are not an afterthought, which means, you can do simple things that all your users wanted, like putting a different image on each desktop, by design, not a crude tacked on hack added late in the release series, like happened with 4.x. Again, being unable to do this I believe is a design flaw, not a small thing that doesn't matter.

2. switching between heavily loaded virtual desktops happens almost instantly. 4x would take up to 15 seconds on a very heavily loaded desktop switch, it was really getting silly. xfce of course manages this, magically, as it should, instantly. So did kde 3.5.

3. multiple monitor support is a core feature, face up to it, only power users as a rule use gnu/linux, it will never crack the windows market, this is particularly true because the one thing regular users not only hate, but won't tolerate, is incomprehensible failures of their desktops. I was a significant kde distro guy, and I lost 1 week on the various failures of 4.x to 5x, and I am patient, and able to debug, and try to resolve issues, heck, I'll even file bug reports to devs who are competent enough to take them, but certainly not if they react with defensiveness (always the hallmark of the mediocre I find).

So there's little point in pretending that gnu/linux desktop marketshare is more than the roughly 1.8, and apparently falling, that it is now, which means, get real, stop doing stupid over the top graphics junk and focus on what matters to users, the desktop working, being fast, stable, and reliable. If you can't figure that out, then just stop pretending kde is more than a hobbiest project.

I loved the entire concept behind kde, but it's failing seriously now. I'd say xfce is only a few core features/weaknesses away from being almost everything kde 3.5 was, which is more than 4.x and 5x ever will be given how slow they are. xfce of course has to deal with the radical gtk 3x instability, so I guess they will wait for gtk 3.x stable to be released, then build on that, incrementally moving along, slowly, conservatively.