First let me say that by no means I want to criticize KDE 4 or its developers. Yes, KDE 4 isn't for me, and I suspect I will still prefer KDE 3 over it for a long time, but I can't see anything wrong with the developers deciding to try something new and writing a new piece of software. After all, the old software is still there for everyone to use and bits don't rot, do they? And KDE 3.5.10 was finished and feature-complete, they could not possibly improve on it much further anyway.

KDE 3.5.10 was released just 4 months ago. It can hardly be called old. I said it was finished and feature-complete, which is like saying that yes, in a sense, it is just perfect. That's why I wonder why many major distributions, namely Kubuntu, Mandriva and Fedora, dropped KDE 3 so early. And the others will probably follow suit soon. I just received definite information that KDE 3 will be dropped from Debian squeeze - that's the version after lenny.

I wonder if there is something about finished and feature-complete - that is to say: perfect - software, that the Linux community doesn't like? When a software is actively being worked on, everyone rushed to ship the latest versions, but as soon as a software is finished, it is immediately being dropped. At least that's the case with KDE 3.5.10, and there are probably other examples.

The distributions could have decided to stay with KDE 3.5.10, which is simply an excellent piece of software, no doubt about it, and many people do want to use it. They could have decided to ship *both* KDE 3.5 and KDE 4 in order to make everybody happy. Why did they decide against it?

Just an ignorant question: Is it really so difficult just to keep a certain piece of software, like KDE 3, in the repositories, so that everybody may use it if he wishes so? KDE 3 and KDE 4 are quite different pieces of software, why not offer them both together? What's too much hard work about it?

There are a few examples of old software not actively being worked on anymore which is still shipped by the distributions. For example, twm, the ancient window manager. Why shouldn't software, written once, live on as long as people want to use it, no matter if it is still actively developed? After all, being in development means that something is yet unfinished, whereas it should be the truly finished things that are best. It makes me sad to think that so much software which people worked so hard and long on to make it good, dies a death so premature.

This is partially an innocent question and partially just a rant. Maybe somebody wants to opine on it or explain to me why it would be too much hard work to ship an already perfected piece of software like KDE 3.5.10.

Mepis, Debian, PCLinuxOS, sidux are all distros that are preparing a new release with KDE 3.5 or have just released a new version, so not all distros made the switch. Apart from that, you can get 3.5.10 on Suse, on Arch (KDEmod) and probably on a couple I forget.

So if you want to keep using 3.5, maybe you need to change your distro....

Then, for the future, it is nothing but logical that distros should switch sooner or later. After all, they need to follow development. KDE3 will eventually be obsolete.

I for one applaud the fact that the KDE developers have had the balls to radically break with the past and write KDE4 from scratch. It has lead to a more stable base, cleaner code, extra features and a better platform to develop new apps on features on. In this business, standing still means getting behind. With all due respect, but the Gnome DE is still very much the same it was 5 years ago, and there is not very much that indicates it will be very different 3 years from now.

I am no coder, but I do know that entropy is a serious risk in development. There just comes a moment that you can no longer clean up the mess you have been creating over the years. KDE4 is now based on shiny new code. I am convinced it will be a much more robust platform than a continued KDE3 will ever be able to give you....

So what was on KDE3 that hasent yet made to KDE4 that you miss so much?

I think it's because of the lack of developers.
The KDE teams of the distro have enough work with KDE 4, that there is no time to maintain KDE 3 as well.
But as long as the distributors support an older version with KDE 3 ... it shouldn't be as bad. Me for example, is still suggesting Mandriva 2008.1 for beginners ... when it's no longer supported, KDE 4 has a stadium were it's nearly as easy as KDE 3 for them.
(K)Ubuntu LTS is even better in that way

Regards,
TeaAge

XiniX wrote:So if you want to keep using 3.5, maybe you need to change your distro....

Yes, I already did that, from Kubuntu to Debian. For now I am running KDE 3.5.10 perfectly fine, I am only worried about the future, maybe in 1-2 years I will still not like KDE 4, but because of other software I depend on, I feel that I have to upgrade.

Then, for the future, it is nothing but logical that distros should switch sooner or later. After all, they need to follow development. KDE3 will eventually be obsolete.

As long as there is still a substantial number of people wanting to use it, in spite of the bugs it has, I don't see why it should be obsolete. And I wonder if distros had abandoned KDE 3, too, if KDE 4 had been given a different name in order to indicate that it is a totally different project, an entirely new DE, just adding another choice to the choices of Desktop Environments.

GNUstep hasn't released for almost 2 years, yet it is widely regarded as a thing of the future. KDE 3.5.10 has released 4 months ago and is regarded as obsolete.

I for one applaud the fact that the KDE developers have had the balls to radically break with the past and write KDE4 from scratch. It has lead to a more stable base, cleaner code, extra features and a better platform to develop new apps on features on. In this business, standing still means getting behind. With all due respect, but the Gnome DE is still very much the same it was 5 years ago, and there is not very much that indicates it will be very different 3 years from now.

It has to be noted, however, that the GNOME DE is very successful, more successful than KDE, 5 out of 6 distributions ship it as the default choice. It seems that many people just want a functional DE and don't care about recent innovations like compositing, screenlets and such.

I am no coder, but I do know that entropy is a serious risk in development. There just comes a moment that you can no longer clean up the mess you have been creating over the years. KDE4 is now based on shiny new code. I am convinced it will be a much more robust platform than a continued KDE3 will ever be able to give you....

This may eventually come true. However, I still don't understand why it would be necessary to continue working on the KDE 3 code at all. Can't we leave it as it is and just use it? Certainly KDE 3.5.10 still has bugs, but many people might want to use it in spite of that. And old, mature software usually contains fewer bugs that brand-new software under heavy development.

I don't understand why a piece of software, once matured, couldn't remain in use indefinitely. I see only three reasons why a piece of software becomes obsolete:

1. because dependencies (e.g. libraries) aren't provided anymore
2. because all users choose not to use it anymore
3. because a serious security hole get discovered but won't be fixed in due time

I understand that the third reason would apply for KDE 3.5.10, but can't we just wait until that actually occurs?

The primary reason would be that the distributions have to support it for the life of the distribution it is shipped with, meaning they have to provide support and ensure compatability with updates to dependencies, which will eventually become a problem.

melkart wrote:I understand that the third reason would apply for KDE 3.5.10, but can't we just wait until that actually occurs?

Sure, just keep using Debian Lenny for its life cycle, which will surely be some 3-4 years or so. Or Kubuntu LTS, which also has some 2.5 years going for it.

Seriously, I do not see the problem. You can keep using KDE 3.5 for at least 3 years or so with full support.

Oh, and as for the Gnome.... it is not as successful as windows, but I am sure we all agree that quantity does not imply quality...

XiniX wrote:Sure, just keep using Debian Lenny for its life cycle, which will surely be some 3-4 years or so. Or Kubuntu LTS, which also has some 2.5 years going for it.

Just a minor correction, Kubuntu Hardy is not LTS. The last Kubuntu LTS is Dapper, for which desktop support ends this coming year.

melkart wrote:It has to be noted, however, that the GNOME DE is very successful, more successful than KDE, 5 out of 6 distributions ship it as the default choice. It seems that many people just want a functional DE and don't care about recent innovations like compositing, screenlets and such.

Well, first: it's a bit risky to "play" the numbers-game. It's hard enough to gauge how many Linux desktops that are currently in use, never mind gauging which DE is used on those.

Secondly, which one is most successful depends on where in the world you live. From what I've seen, Gnome seems more popular in the Americas (and the U.S. in particular) while KDE is more popular in Europe. In my experience, that certainly seems to be the case of my region: scandinavia. I see more KDE desktops than Gnome ones (albeit with some mixing of the apps, i.e. K3B being used in Gnome and Gnome-do being used in KDE).

And if recent innovations, as you put it, were not deemed important by the users: how come compiz and google-gadgets (as just one example of widgets) are so popular regardless of DE?

Kryten2X4B wrote:And if recent innovations, as you put it, were not deemed important by the users: how come compiz and google-gadgets (as just one example of widgets) are so popular regardless of DE?

I was not aware that Google Gadgets are so popular. I had in mind that the most popular operating system today is still Windows XP (>60% market share) which has none of these recent innovations. It just sports the standard, old-fashioned DE features, but, in my opinion, implements these very well. Some recent innovations might have some merit, but I can't imagine how anyone could possibly find wobbly windows or rotating widgets useful.

I can't imagine how anyone could possibly find wobbly windows or rotating widgets useful.

Not useful but beautiful.
I'm ashamed to say this, but I use KDE 4 because it's more beautiful than KDE 3. Apple User pay a lot of money only for design.

I agree with you, that design isn't everything and that KDE 4 still has some lacks.

Nevertheless, the KDE Teams of the distribution has not enough time to support KDE 3 and KDE 4 in there distribution.

Regards,
TeaAge

melkart wrote:but I can't imagine how anyone could possibly find wobbly windows or rotating widgets useful.

Maybe not useful, but people seem to like it.

TeaAge wrote:Apple User pay a lot of money only for design.

Yes, but for the near perfect usability and integration, not for eye-candy. This is where Linux DEs are sorely lacking, GNOME a bit less than KDE 3. About the usability KDE 4 I have nothing to say.

Nevertheless, the KDE Teams of the distribution has not enough time to support KDE 3 and KDE 4 in there distribution.

But they do have the time to support a lot of less important stuff. For instance, Debian also supports Xfce, LXDE and many other window managers. I am pretty certain that, by today, still more people want to use KDE 3 than Xfce or LXDE. Why did they drop KDE 3 for squeeze, but keep Xfce and LXDE then?

The fact that Debian lenny will still be supported for a long time, does not help the many people who want or even need to use Debian testing or sid. Once lenny is released, KDE 3 will be dropped from testing. If you want to continue using KDE 3 with Debian you will have to do with a lot of other outdated software.

Hans wrote:

melkart wrote:but I can't imagine how anyone could possibly find wobbly windows or rotating widgets useful.

Maybe not useful, but people seem to like it.

Those people voting at kde-look.org seem to like it, and they are only a tiny subset of the user base. I can't imagine people getting excited about wobbly windows when they want to use their desktop for work.

admoore wrote:Just a minor correction, Kubuntu Hardy is not LTS. The last Kubuntu LTS is Dapper, for which desktop support ends this coming year.

I stand corrected, you are right of course....

OK, Debian Lenny, PCLOS and MEPIS will be some options for a KDE3 implementation with some longevity....

melkart wrote:I was not aware that Google Gadgets are so popular. I had in mind that the most popular operating system today is still Windows XP (>60% market share) which has none of these recent innovations. It just sports the standard, old-fashioned DE features, but, in my opinion, implements these very well. Some recent innovations might have some merit, but I can't imagine how anyone could possibly find wobbly windows or rotating widgets useful.

I was mostly thinking about the Linux desktops to be honest, but sure. XP is old-fashioned if you will (and when I have to use Windows, I sure hope XP is the one installed. It's the least annoying windows variety). But yeah, I've seen quite a lot of XP installs where the owner has installed things like WindowBlinds, Google gadgets/desktop x/yahoo's gadget system, and what not. Necessary? Maybe not, but can add useful features. And sometimes whether it is useful or not does not even enter the picture - what is sometimes wanted is merely to be able to get your computer to behave and look like YOU want it to. Which coincidentally can be a productivity booster in its own right.

Wobbly windows may be a bit silly, but there are other useful features of composite (such as present windows and the desktop grid). Rotating the widgets well, I consider that a "side effect" of the widgets being able to adjust to the size of where it is placed (desktop vs. panel for example). If it is necessary for it to be able to scale, it may not (note: I'm not a programmer so I am only extrapolating here) be much harder to make it possible to rotate them as well.