Phoronix News Bot
KDE's KWin Gears Up To Advance At Faster Pace
Phoronix: KDE's KWin Gears Up To Advance At Faster Pace
With the push towards Wayland and other various advancements being desired out of KDE's compositing window manager, Martin Gräßlin is joining Blue Systems to further work on KWin...
http://www.phoronix.com/vr.php?view=MTIwMDg
Senior Member
Oh, how wrong people were dismissing KDE when 4.0 came out. Its evolution was just short of amazing ever since. I can only hope Gnome3 will follow suit, though imho its current incarnation is unusable for desktops. Then again, maybe I'm getting older and not adjusting to change as I once did.
Senior Member
People dismissed KDE 4.0 because it was beta or alpha quality when it came out...what people didnt notice was that it was tagged KDE 4.0 Beta / Alpha in the code. It was only supposed to be a snapshot release "Hey this is the idea we're thinking! DONT FREAKING RUN IT ON PRODUCTION!" But Distros picked it up as "Stable" and so everyone ran it, everyone got crashes, bugs, segfaults, etc and everyone got mad about it. What ended up happening was 4.0 ended up becoming "alpha", 4.1 "beta" and then 4.2 was when things stabilized out and started to become more of an evolutionary iteration getting better and better.Originally Posted by bug77
KDE 4.9.2 on my desktop and loving it aside from 1 or 2 things
Senior Member
That simply isn't true, many people attacked KDE for its fundamental design decisions in KDE 4. Lots of people were attacking plasma for being too different than traditional desktops, attacking kwin for not using compiz for compositing, attacking dolphin for not having all the features of konqueror, attacking KDE for using phonon and not gstreamer exclusively, and attacking KDE 4 in general for relying too much on abstraction layers, amongst other things. However, all the fundamental design decisions that KDE developers made, but were heavily criticized for at the time, proved to be the correct ones.
Phoronix Member
@TheBlackCat You are right. In fact, the best technical decision was Solid. Since KDE was introduced before anyone considered dropping HAL and friends, just imagine what would have happened if KDE used HAL directly.
How right or wrong KDE decisions are is proved, in fact, by an objective measure: how many forks there are that rely on the criticism made to a DE, and how popular they are. I don't know about a main distro using Trinity, but one of the most popular ones formerly using GNOME, Mint, now uses such a fork: Cinnamon.
Last edited by Alejandro Nova; 10-07-2012 at 05:37 PM.
Senior Member
I find gnome 3 to be a very usable desktop. KDE is nice too, but I still tend to prefer gnome these days. KDE sure has come a long way since 4.0 though.
Senior Member
KDE is chasing after new features rather than fixing the thousands of bugs it has right now. This is one of the reasons Gnome and Unity will remain the more popular choices.
What KDE needs right now is a year of no new features but bugfixes and stabilization. That won't happen, of course. Someone else said in some blog that KDE has so many bugs because "the fun is over; now it's hard work." I totally agree with that now. Chasing down bugs is boring, so no one fixes them.
KDE has bitten off more than it can chew. All the features sound like a good idea. But when nobody wants to maintain them, then it's not such a good idea anymore.
Senior Member
uh-huh, especially amazing is how it's 4.9.2 now, and it's still doesn't have feature-parity with 3.X in multi-monitor department.
and even then it wasn't ideal, now, with things like KMS, DMA-BUF, "DRI3", Wayland and X-independent Mesa acceleration techniques, there are even more possibilities... but even 5-10 year old ones are not properly supported.
ah, and don't get me started on slow-ass useless glitchy ugly crap of nepomuk/akonadi with their MySQL requirement and idea of file-indexers in general (i already have logical file order in a file-system, fuck off from my HDD and CPU, please !).
at least now they starting to use things like Telepathy and Gstreamer. guess someone tired of fucking around.
indeed !
even now, when i updated from 4.9 to 4.9.2 i got instability issues as regressions, like khtml being broken, and everything depending on it going apeshit, if being compiled with the same flags as on 4.9.
BUT things like that, no matter how frustrating, still are the LEAST of KDE's problems. in comparison to fucked-up design and lack of manpower/interest in crucial parts, like kwin.
Junior Member
Where did you read that? I'd be curious to read it.
I agree that KDE team had a too ambitious project.
I, for one kept distant from kde, as the DE was too unstable compared to Gnome (even 3.0, despite the fact that people might like it or not, Gnome 3.x never crashed on me).
But right now I'm on kde 4.9.2 and it seems very stable, much more stable than in the past (even compared to 4.9.1 or 4.9.0).
I am under the impression than in 4.9 and 4.8 releases they added relatively few features and tried to stabilize more the DE.
And reading here and there, it seems that they'll keep doing it for 4.10 release too (I read about Martin Groblin's blog about optimizing performance for 4.9.2 and 4.10, as also read the new about fixing the problems with AMD drivers..).
So while in the past what you said might be true, it seems to me you're wrong (or at least not totally fair) about the present.
As Martin Groblin wrote, often developers do not like to talk about optimizing performances or fixing bugs, because that means they did something wrong when initially writing the code..
Phoronix Member
I always thought that too (plus maybe also including performance improvements).
Or rather an entire major release (4.11?) in which the only accepted features would be strictly the ones in the top 10 most voted by the community (which I guess are the ones that bring feature parity with 3.x).
No more "let's redesign this dialog for 9nth time"-type of "features", please.
Reply With Quote
