KDE 4.10 Desktop Delayed Into February

Quote:

---

Originally Posted by zanny

I agree with making KIO a fuse fs. I am a recent KDE convert, but one of my bigger issues with the project is the consistent ideology of "if it exists outside of the qt / kde environment, reinvent the wheel and rebuild it". I agree with the overall stance that as a development ecosystem qt / C++ / QML is drastically superior to Gnome / GTK / C, but it really divides the Linux desktop with all their personal reimplementations of core features.

---

What specific examples are you referring to? For most of the examples I have seen people list the KDE version actually came first.

For example, as far as I have been able to determine KIO predates FUSE by at least 3 years. Similarly KOffice predates openOffice.org, konqueror predates firefox, khtml predates gecko (and webkit is a fork of khtml), and nepomuk predates tracker.

The only major example I am aware of where KDE re-implemented something that was already available was with kwin's compositor vs. compiz, and all the problems compiz faced until Ubuntu took them over demonstrated that was a good decision. Of course there may be some third-party projects that created a KDE implementation of an existing piece of software, but it doesn't appear to be any more common than with other toolkits.

Quote:

---

Originally Posted by Awesomeness

So LightDM-KDE does not exist?

---

I meant KWin, not KDM. The display manager has nothing to do with the graphics stack of choice. I don't even use KDM actually, I login at console and have xinit in my bashrc.

Quote:

---

And as a recent “convert” you suddenly obtained all knowledge about KDE?
Your claim is absolutely false and I can easily prove it: https://projects.kde.org/oxygen-gtk

---

Actually, yes! I downloaded the KDE suite. Into my brain. It's like I'm Neo from the Matrix over here, it all seems so real.

And I do acknowledge they have gotten a lot better about using tools with platform momentum in recent times, be it pulseaudio, dbus, or gstreamer. And oxygen-gtk is just a gtk port of the default theme, it isn't major systemic infrastructure like what I meant, and of course there are exceptions, its open source. I'm talking about how Phonon and Gstreamer ran parallel for a while, and are now mingling. I can get behind the sentiment of something like "I like gnome-system-monitor, but our desktop doesn't have GTK dependencies, so we will write our own in qt, and expect it to be better for our choice of toolkit, so heres KGuard".

Quote:

---

KDE’s current approach is to either integrate GTK applications as well as possible via Oxygen-GTK or to write a KDE front-end on an existing back-end (as with LightDM, NetworkManager, PackageKit, etc.)
In fact the main reason why there is no broad push for a KDE web browser (Konqueror and Rekonq are basically one-man shows these days) is that most KDE devs simply use Firefox or Chrome.

---

Oxygen-GTK is just a gtk theme that mimics Oxygen stock. As soon as you start changing themes, window decorators, and widget styles, you need to tweak the gtk themes. That isn't kdes fault at all though that gtk apps use their own themes, of course they do.

And I actually liked the ideas behind Konqueror as a universal resource explorer. I do think it is a good thing projects like Rekonq are downplayed because the best way to make KDE better is to focus on the aspects of the experience that don't have tremendous market competitors where you try to tuffle with giants. It is like how Totem is mediocre next to VLC because it is just another part of Gnome vs VideoLan being entirely centered around VLC.

Quote:

---

That's because the current keyboard shortcut implementation is lacking and IIRC cannot easily be fixed without breaking binary compatibility. I believe it’s on the agenda to be fixed in KF5. The current implementation can’t handle two alternatives for global shortcuts. Therefore a default that works also with keyboards without multimedia keys is chosen.
Again: Distributors can change such defaults to target a specialized user group (e.g. one that has a higher likelihood of owning such keyboards).

---

I'm not saying it is a flaw, just an experience. And good to hear it is being looked into, if only for the fact that people expect it since Gnome and Windows provide it (the two places you attract a userbase from).

Quote:

---

Originally Posted by zanny

I meant KWin

---

They have already stated that KWin is written well enough that a rewrite would be pointless. Porting is all it needs.

Quote:

---

Originally Posted by zanny

I'm talking about how Phonon and Gstreamer ran parallel for a while, and are now mingling.

---

Phonon was never anything like Gstreamer, it is a wrapper that allows KDE apps to support multiple media backends, including gstreamer and vlc, and allows KDE to avoid binary compatibility breaks (something KDE guarantees but Gnome doesn't). It doesn't replicate any gstreamer functionality and never has. And it has supported gstreamer all along, despite the gstreamer developers' early promises to support Phonon and then not doing so.

Quote:

---

Originally Posted by zanny

I can get behind the sentiment of something like "I like gnome-system-monitor, but our desktop doesn't have GTK dependencies, so we will write our own in qt, and expect it to be better for our choice of toolkit, so heres KGuard"

---

Again, ksysguard predates gnome-system-monitor by 2 years, and even it was a rewrite of an older KDE system monitor. This was the point I was trying to make. Just because KDE offers something that somebody else also offers does not automatically mean the other version came first.

Quote:

---

Originally Posted by zanny

That isn't kdes fault at all though that gtk apps use their own themes, of course they do.

---

Qt (and by extension KDE) has no problem using gtk themes, but gtk refuses to support Qt themes.

Quote:

---

Originally Posted by TheBlackCat

Phonon was never anything like Gstreamer, it is a wrapper that allows KDE apps to support multiple media backends, including gstreamer and vlc, and allows KDE to avoid binary compatibility breaks (something KDE guarantees but Gnome doesn't). It doesn't replicate any gstreamer functionality and never has. And it has supported gstreamer all along, despite the gstreamer developers' early promises to support Phonon and then not doing so.


Again, ksysguard predates gnome-system-monitor by 2 years, and even it was a rewrite of an older KDE system monitor. This was the point I was trying to make. Just because KDE offers something that somebody else also offers does not automatically mean the other version came first.


Qt (and by extension KDE) has no problem using gtk themes, but gtk refuses to support Qt themes.

---

Also it is necessary to add that, GTK and GNOME were rewrites of Qt and KDE respectively in the first place.

edit: GTK was GIMP Tool Kit originally, but later it became a rewrite of Qt practically, because Qt license wasnt "free enough". Even when Qt become GPL or LGPL the hatred continued.

Oh man. How come every single KDE related thread begins with a flamewar ? Please cut us some slack ! (not sure it's proper English :)

Is the Phoronix community made of 15 years-old with much time to lose people ? (apparently, my feeling is just due to a couple of trolls but that's unpleasant :)

As a long term Linux user, spending SO MUCH TIME trying every single distro & environment, well, I like KDE, Unity, Cinnamon, Gnome Shell, XFCE & many other lighter environments. But, it's so frustrating, overall. Each time I change, I find so many flaws that weren't in my previous environment. KDE is so brilliant to some extent, it sometimes seems wasted with unfinished features, and it seems to be "stabilizing" forever. But whenever I switch back to Unity/Gnome&co, I have the feeling I made a big jump into the past for some aspects (and in the future for some others). Also, even it feels there are always new bugs in KDE, there are also some very long standing ones in each environment.

So, well, I wish there were fewer & better choices. Each environment is great to some extent, and lacks much too. How depressing : if only we could have to best of each of them ! ;-)

Quote:

---

Originally Posted by zanny

I agree with making KIO a fuse fs. I am a recent KDE convert, but one of my bigger issues with the project is the consistent ideology of "if it exists outside of the qt / kde environment, reinvent the wheel and rebuild it".

---

I think that this is FUD that people have started accepting.

Most of the things that people claim KDE "reinvented" were in fact first introduced by KDE, and then copied by everyone else. When KIOslaves were introduced, there was nothing comparable. KOffice was older than OpenOffice. KHTML existed back when there were no good open source browsers (Netscape was closed). There was nothing like DCOP (CORBA was way too heavy for a desktop back then), which became DBUS. KDE even did "menus on top" like Unity, and did it 16 years ago. Karamba provided desktop widgets for KDE before anyone else had them.

It is usually a case of the other desktops not wanting to use KDE stuff and then inventing their own. DCOP was turned into a desktop-agnostic DBUS, which is the way it should be done, IMHO. KHTML became Webkit, which benefitted everybody. But akonadi (no KDE deps) was designed as a desktop-neutral standard, but rejected by GNOME. GNOME attempted CORBA for a decade before finally giving up on it and going with a simpler system (DBUS). PulseAudio was developed from scratch, instead of reusing aRts or ESD. Or Jack.

I don't think that KDE is the major NIH culprit, like you seem to think.

Quote:

---

Originally Posted by zanny

I meant KWin, not KDM.

---

This makes sense, actually.

KWin is a really mature window manager. It is simply BETTER than most other window managers, because of decades of logic. It is far better to use on a daily basis than Compiz, although the efects may not be as funky. I used Compiz with KDE for a few years, and when I switched back to KWin, things became so much better all of a sudden.

Quote:

---

And I do acknowledge they have gotten a lot better about using tools with platform momentum in recent times, be it pulseaudio, dbus, or gstreamer.

---

I think that you are confused. DBUS is a desktop-agnostic version of DCOP, a KDE technology. They work exactly the same way, with exactly the same command line options and syntax. In essence, every desktop adopted KDE technology when they switched to DBUS. KDE didn't have to change anything.

GStreamer is a backend. KDE could always use multiple backends, including GStreamer. Even today, you can choose between VLC, Xine (deprecated) or GStreamer for decoding.

PulseAudio is supported for people who are forced to use it. It is not a part of the KDE desktop, nor is it necessary.

Quote:

---

And oxygen-gtk is just a gtk port of the default theme, it isn't major systemic infrastructure like what I meant

---

See, the thing is -- if you start a Qt applications inside a GNOME session, it will automatically change its appearance to match the GTK theme and integrate perfectly. The Qt guys did this to make the GNOME users' life better.

The GTK guys refuse to do the same, although code has been out there for almost a decade. So KDE HAS to provide themes like that one.

Same thing happened with Firefox and LibreOffice. There were Qt front-ends for both of them, making them fit in perfectly with the KDE desktop (like they do with GNOME). Unfortunately, most Firefox and LibreOffice coders don't understand Qt, so it was not maintained and the effort died.

Quote:

---

And I actually liked the ideas behind Konqueror as a universal resource explorer. I do think it is a good thing projects like Rekonq are downplayed because the best way to make KDE better is to focus on the aspects of the experience that don't have tremendous market competitors where you try to tuffle with giants. It is like how Totem is mediocre next to VLC because it is just another part of Gnome vs VideoLan being entirely centered around VLC.

---

The important thing is apps like Firefox of LibreOffice or VLC seem to work better in GNOME because they either chose to ignore Qt front-ends (Firefox and LibreOffice) and use only GTK, although neither of them is a GTK app, or because best Qt apps integrate well with GNOME (VLC, skype, Clementine, anki...) so people don't notice that they are using separate apps and think that they are a part of the desktop.

On the other hand, GTK apps do their best to stick out like a sore thumb.

If GTK finally merged Qt support, we wouldn't be having this discussion. But their primary purpose has always been killing KDE, not friendly competition.

Quote:

---

Originally Posted by pingufunkybeat

like Firefox of LibreOffice or VLC seem to work better in GNOME because they either chose to ignore Qt front-ends (Firefox and LibreOffice) and use only GTK, although neither of them is a GTK app, or because best Qt apps integrate well with GNOME (VLC, skype, Clementine, anki...) so people don't notice that they are using separate apps and think that they are a part of the desktop.

On the other hand, GTK apps do their best to stick out like a sore thumb.

If GTK finally merged Qt support, we wouldn't be having this discussion. But their primary purpose has always been killing KDE, not friendly competition.

---

Qt only does what it states in its mission statement : write once, run everywhere (yeah I know there are some problems...). The point being that after adding Cocoa (OSX) and windows themes support to look as native as possible, supporting GTK themes was easy.
AFAIK that "primary purpose" is gone, or at least fueled mainly by users.
The rejection of akonadi deeply saddens me, because I see it as an awesome framework opening so many possibilities, but there lies the greatest weakness of KDE : so many awesome ideas and so few people to implement them. (few in a way relative to the scope of the project). Calligra is finally usable, with most apps (the only exception I know of being Krita) being lead by one or two people, thats a major milestone.

Quote:

---

Originally Posted by Serafean

The rejection of akonadi deeply saddens me, because I see it as an awesome framework opening so many possibilities,

---

Hu? When got akonadi rejected?
It has problems, for sure. I completely wiped all nepomuk+akonadi from my disk, as it caused headache to me - but many people are happy with it, and it also seems to work for them.

Quote:

---

Calligra is finally usable, with most apps (the only exception I know of being Krita) being lead by one or two people, thats a major milestone.

---

krita was the only usable app for quite some time.
sheets and words had major compatibility issues with open document. crashed alot. But krita was usable all the time. And actually many artists use it, so if krita was the exception, then just because it is excellent ;)