KDE Plasma 5 Officially Released

[Sho_]

Does the KWallet integration with KDE 4 apps work with Plasma 5 (seems broken here) or will it only work with KF5 apps?

TBH I actually don't know, I have to defer to folks who worked on KWallet bits there - try catching apachelogger on Freenode e.g.

(I work on core Plasma Desktop bits and a bunch of apps.)

[Luke_Wolf]

Both Dolphin and Konsole are already ported in git, BTW - many of us have been dogfooding those ports daily for weeks. They're just not quite release-ready yet, and we're staggering releases rather than trying to do it all at once (remember how that went with KDE 4 ). Keep in mind this release and news post is about Plasma 5, and those applications aren't part of Plasma.

And that you can keep using KDE 4 apps in Plasma 5 is a good thing, not a bad thing .

Does the KWallet integration with KDE 4 apps work with Plasma 5 (seems broken here) or will it only work with KF5 apps?

[Sho_]

Both Dolphin and Konsole are already ported in git, BTW - many of us have been dogfooding those ports daily for weeks. They're just not quite release-ready yet, and we're staggering releases rather than trying to do it all at once (remember how that went with KDE 4 ). Keep in mind this release and news post is about Plasma 5, and those applications aren't part of Plasma.

And that you can keep using KDE 4 apps in Plasma 5 is a good thing, not a bad thing .

[mark45]

Impressions:

1) The live (neon) iso actually booted up on nouveau. Good. The iso is i686 btw.
2) All/most apps are qt4/kde4 (dolphin, konsole)
3) "sudo apt-get update" and then upgrade broke the system, doesn't boot.
4) Little to see since (1) all apps are Qt4/kde4, (2) Plasma 5 is almost all about under the hood stuff, (3) the new theme and icons are yet to be polished/finished, (4) dolphin and all other core gui apps aren't ported to Qt5.
5) (Full) Wayland support will come in the future.
6) The boot up (before the update which borked the system) was rather slow.

To sum it up: Don't get too excited, the desktop is raw, needs another 3 years to live up to its potential (including Wayland support by default).

[Sho_]

if so, how much code was ported to QML?

99% of Plasma 5 is QML on Qt Quick 2 (and thus rendered through OpenGL), including settings UIs, which use Qt Quick Controls now. On platforms that support it properly we now have dedicated render threads, animations running on background threads, and dedicated render threads per window.

We still have more improvements to make to use OpenGL even more efficiently (some of our low-level components don't do everything they could yet to reduce CPU usage, texture uploads and video memory usage), though.

[workless]

animations seems to be well done, for example widgets borders and clock widget, were they ported to QML2?
if so, how much code was ported to QML?
is there any plan to port settings application and other to QML step by step?

[cfeck]

You know what would be awesome? Instead of total icon redesigns, every project lead (or whatever KDE calls them) could do their app icons however they want, then have a themeing program that modifies all the icons based on the theme specs, whether it's monochrome, oxygenlike, pastel, etc. That way, only a single change system would have to be changed with the themes, instead of every single icon everywhere on the desktop. I know it's probably harder to make than it sounds, but it would theoretically solve all of the icon theme issues.

That is an artificial intelligence problem. Good luck.

[Luke_Wolf]

So I decided to try out Plasma 5 on openSUSE and the KDE SC 4.x Applications seem to have broken KWallet integration (in particular Kontact and KTP), also kicker locks up plasmashell (I thought that a badly behaved plasmoid crashing/locking plasma was supposed to be solved through QML?) and secondlife crashes kwin and plasmashell which restart fine, but eventually the desktop locks up such that only the cursor responds clicks don't work and keyboard is nonresponsive.

The project neon image works fine though as far as the kicker goes, didn't test KWallet or secondlife, anybody else having these issues (want to be sure it's not just me before I file bugs)?

[Prescience500]

You know what would be awesome? Instead of total icon redesigns, every project lead (or whatever KDE calls them) could do their app icons however they want, then have a themeing program that modifies all the icons based on the theme specs, whether it's monochrome, oxygenlike, pastel, etc. That way, only a single change system would have to be changed with the themes, instead of every single icon everywhere on the desktop. I know it's probably harder to make than it sounds, but it would theoretically solve all of the icon theme issues.

[Sho_]

Awesome, I'll definitely give it a go then as soon as it appears on the official arch repositories. It does look like a huge step forward, but as the posters above said the default KDE interface always seems to look ugly. I've always felt that KDE would get a lot more interest/market share if it shipped with a better default theme. That said, (from the video) it looks like Plasma 5 is a step in the right direction.

The problem is getting from "ship with a better default theme" to actual assets. It's not that "I don't like it" isn't useful feedback, but it's (a) not very specific (and you need to make specific changes in the end) and (b) it doesn't by itself tell you anything about the representation of that opinion across the entire audience. You can also assume that the people making the assets we ship don't intentionally make something sub-par, and are generally critical of their own work.

There's also some bigger picture things at work, though: KDE as a development community has a certain set of ways-of-working that have served us quite well, such as the way we manage developer recruitment and write access, share responsibilities, and so on. KDE has been around for almost 20 years now, grown significantly in size in that time, shipped regularly, and is now in roughly its fourth developer generation. Many other FOSS communities aren't nearly as turnover-proof or scalable. The challenge we face is how to balance work that is somewhat easier to accomplish with shorter leashes and more secrecy (such as implementing visual design) with these these things that have allowed us to continue to exist, without compromising either. I'm quite confident we're making progress on how to solve these things, just as we've worked through others in the past. If anything, our stability gives us more time to figure it out: We're intent on being around in another 20 years.

It's always a good idea to assume the people doing something aren't mindless, do care about what they're doing, and have to balance concerns .