2016 Wayland Experiences: GNOME: Perfect, KDE: Bad, Enlightenment: Good

[smitty3268]

I'm not a native speaker, but I disagree with you. In my opinion, most of the time "perfect" doesn't mean "faultless" but "excellent in all aspects" or even "beyond expectations".

Why bring opinion into it? Just look at the dictionary definition:

1

a : being entirely without fault or defect : flawless <a perfect diamond>

b : satisfying all requirements : accurate

c : corresponding to an ideal standard or abstract concept <a perfect gentleman>

d : faithfully reproducing the original; specifically : letter-perfect

e : legally valid

Perfect does mean "faultless". If you want to say "beyond expectations", then say that instead.

Edit: Ok, so maybe it doesn't have to be 100% flawless - if you get detailed enough perhaps you can find a fault while still originally saying it was perfect.

But my point is that you can't specifically go out of your way to point out multiple flaws, and then in the next sentence claim it was perfect. That just makes no logical sense. If it's perfect those flaws are beyond notice.

[Awesomeness]

I wonder who the KDE guys will blame now?

They cant blame Wayland, since that is "production ready since ages". I bet they blame Canonical, as always, because Canonical made MIR and the KDE guys were too busy flaming about it and "proving" how much better Wayland is.

Plasma 5.6 will be the first release even with desktop Wayland in mind. Earlier releases were only a byproduct of Plasma Mobile.

[DarkCloud]

perfect: make (something) completely free from faults or defects, or as close to such a condition as possible.

As for the whole Gnome vs KDE nonsense, Gnome kinda had a major head start. Like, over a year head start. It seems like most people forget KDE was in the middle of their gigantic refactoring of code going from &quot;KDE libs&quot; to &quot;KDE Frameworks&quot; when Gnome started on Wayland. The KDE devs didn't even officially start Wayland work until what? less than half a year ago? Yet it runs on Wayland, even if it's not a great experience. All this while they're still working out bugs from the major shift to the KF structure and other apps being ported to it.

I'd call that pretty good progress if I do say so myself.

But kde is always in the middle of a refactoring. I used to have Kubuntu on all my computers, but now just one of my three computers has it and it is a shared partion with Windows 10, and sadly Win10 gets more bootups. I am happy with Unity, it doesn't do as much as kde, but that's kind of the beauty in it ,as it just works and doesn't get in the way.

[Delgarde]

As for the whole Gnome vs KDE nonsense, Gnome kinda had a major head start. Like, over a year head start. It seems like most people forget KDE was in the middle of their gigantic refactoring of code going from "KDE libs" to "KDE Frameworks" when Gnome started on Wayland.

So? Gnome had to do their big refactor too - support for Wayland was one of the motivations for all the big API breaks... not the only one, but reducing the hardcoded dependency on X11 was a big part of that effort. Both projects have had a similar amount of work to do, to take something built around the assumption of X11, and evolve into something that allowed for other possibilities.

[ssokolow]

Trolling aside, KDE until this point has been very explicit about Wayland still being a Work-in-Progress, so I'm not sure how you draw this equivalence.

That said... I like KDE too much to even try GNOME/Wayland even though I'm really itching to have an X11-free desktop. There's no doubt that Wayland is the future on Linux (and hopefully also the BSDs!), but I'm actually quite contented to wait until my compositor is stable enough. We have x11-wayland for a reason.

No argument on the GNOME front. I've always been a huge fan of sufficiently stable and snappy KDE versions (eg. the 3.5.x series, current 4.x, etc.) and I'm more likely to assemble my own desktop from various less-than-easy-to-setup components than to feel comfortable on GNOME.

I'm referring to how the people I talked to justified taking so long to get things responsive and stable after 4.0. (I finally gave up on KDE 4.x and switched to LXDE around 4.6, only returning to KDE maybe six months ago.)

Specifically, they pointed out that KDE 3.x and 2.x followed a similar pattern and gave a strong impression that doing so somehow justified not learning from the experience and repeating the pattern with 4.x.

Even now, Kwin 4.11.11 gets hung up and requires a Ctrl+Alt+F1 (because compositing has frozen up) and a kill/restart every week or so. Having that kill my login session and anything unsaved at the time would be flat-out unacceptable.

[molecule-eye]

I really hope KDE Wayland gets into usable shape in the not so distance future. I tried Gnome 3.18.x on Wayland on my Ivybridge hardware on Ubuntu Gnome 16.04, and the cursor was laggy and jerky which made it unusable. So it is definitely far from perfect unless Intel is entirely to blame.

[fabdiznec]

I'm referring to how the people I talked to justified taking so long to get things responsive and stable after 4.0. (I finally gave up on KDE 4.x and switched to LXDE around 4.6, only returning to KDE maybe six months ago.)

Yeah, the whole 4.x thing was badly botched PR-wise. They didn't take into account the absurd optimism of many users... who just plunged ahead, upgraded and ... WTF?

Even now, Kwin 4.11.11 gets hung up and requires a Ctrl+Alt+F1 (because compositing has frozen up) and a kill/restart every week or so. Having that kill my login session and anything unsaved at the time would be flat-out unacceptable.

Oh, that's bad. I assume you've reported it as a bug?

(I had similar issues when upgrading KDE/-frameworks/-whatever on my Arch Linux. Ended up having to kill my login session and restart. Thankly, I'm so used to the Ctrl-S (or ^X^s in emacs) invocation that no work was lost. That's not an excuse for the whole "upgrade K* makes your current running screen locker go weird" issue.)

[Guest]

I wonder who the KDE guys will blame now?

They cant blame Wayland, since that is "production ready since ages". I bet they blame Canonical, as always, because Canonical made MIR and the KDE guys were too busy flaming about it and "proving" how much better Wayland is.

Gnome devs usually blame users. KDE devs don't need to hurry in supporting Wayland. Gnome will be a proving ground.

[mgraesslin]

I wonder who the KDE guys will blame now?

They cant blame Wayland, since that is "production ready since ages". I bet they blame Canonical, as always, because Canonical made MIR and the KDE guys were too busy flaming about it and "proving" how much better Wayland is.

We don't have to blame anyone. We never said it's ready to use. If people want to try it and conclude that it's not ready yet: totally fine.

What's not fine is trolling everything KDE related and bringing up Canonical and MIR all the time as you do. Honest question: don't you have better things to do in your life?

[CrystalGamma]

GNOME on Wayland is perfect except:
* No relative mouse cursor: needed in games (is this a protocol thing?)
* No drawing tablet support (there is no subprotocol yet, about a year after support being made for libinput IIRC)
* gnome-shell is buggy as shit – extensions blocking their thread can freeze the full desktop (even the mouse in wayland; this happens mostly when I use an unresponsive FUSE filesystem for my backups, so my free space monitor will block when it's busy) and I get crashes on average once a week

So there is two things the spec makers still have to do, and one thing the gnome guys need to do.