KDE Commit-Digest for 14th October 2007

In this week's KDE Commit-Digest: Replacement of the "toolbox", and a new KRunner in Plasma, with many applets moving from playground into extragear in KDE SVN. SuperKaramba is now fully integrated into Plasma. A move away from KDEPrint facilities, towards more basic functionality for KDE 4.0. More work on restoring the functionality of the Klipper applet. Basic sound support in Parley. General work on KHTML, with more specific work on image loading and testcases. More work on KDE colour scheme handling. A rethink of device handling for Amarok 2.0. Generic style saving work in KOffice. Various optimisation work across KDE. Kaider moves to extragear-sdk. Kicker is removed from KDE SVN. KDE 3.5.8 and KDE 4.0 Beta 3 (with Kickoff included as the menu option) have been tagged for release.

Dot Categories: 

Comments

by Christian (not verified)

Fist off, just wow. What an impressive list of stuff and an amazing amount of work.

Anyway I noticed a lot of commits about khtml. It all sounds really cool, but can anyone explain what will become of this work once trolltech releases a version of qt with webkit? If KDE moves to webkit (which sounds like it would make sense to the more developed thing) might it be possible to port these changes over? Will work be dropped? Is webkit already using similar stuff?

Anyway super job. Aside from the complaints about qprinter, it sounds like things are really going to rock, even for a 4.0 release.

by SadEagle (not verified)

There are presently no plans in the KHTML development community to abandon development of KHTML and the KHTML library. There is tons of cooperation w/Apple on the JS side of things; and the codebases are very similar, but KJS development is still independent. On the HTML side of things, no one has ever offered an integration path that makes sense, that doesn't throw out tons of advancements, and that does not compromize project values by putting volunteer developers into 2nd class status.

by Emil Sedgh (not verified)

Im really glad to see the khtml is still in development.as a web developer I just want to say kudos, khtml rendering is so nice, better than Gecko in most cases and sure its so faster.

only things that i think are needed in Konqueror as a web browser are:
1)WYSIWYG mode.as i looked around there was a kafka project which is dead for about 4-5 years...is there any plan to implement it?

2)Better javascript debugging...a Web Browser, needs help from webmaster to grow, can you see how 'SpreadFirefox' really worked?and to attract web developers Konqueror needs something like 'Firebug' and 'Web developer toolbar'.
Konqueror's JS debugging looks so weak to me and thats the only reason im keeping firefox: just to see better javascript debugging output

I really tried to do not say my personal needs, these are what i get when talking to other users and designer/developers.

sorry for my bad English

by James Richard Tyrer (not verified)

IIRC: Kafka was renamed Quanta which was renamed KDEWebDev. And so it goes....

by Emil Sedgh (not verified)

iirc kafka was a differrent project, and Quanta is a Part of KDEWebDev module which contains KLinkStatus, Quanta+, Kommander and 1-2 others...

and as far as i know, kdewebdev only contains klinkstatus for kde 4.0 because Quanta+ depends on kdevplatform module which will be ready for 4.1
so no quanta+ for 4.0 :)

by Andras Mantia (not verified)

No, the Kafka code was taken and was improved and integrated into Quanta, where it is called VPL. But unfortunately it didn't really meet the expectation.
As the other poster said, there won't be a Quanta for 4.0.

by Chani (not verified)

better javascript debugging would indeed be very useful.
a while back I had a job that involved working on some rather wacky javascript which was designed to work in both FF and IE. it was utterly unusable in konq, and I wanted to fix that, but I had to give up because I just couldn't debug it.

by Diederik van de... (not verified)

> On the HTML side of things,
> no one has ever offered an integration path that makes sense that doesn't throw
> out tons of advancements, and that does not compromize project values by
> putting volunteer developers into 2nd class status.

Sounds like good arguments. :-)

I'm curious, what integration paths didn't make sense?

by jospoortvliet (not verified)

It would really rock if KHTML and WebKit could merge to a point where they would be (almost) compatible - maybe the biggest problem for a webbrowser is marketshare. Only with marketshare, sites are tested against a browser, and if sites aren't tested against it, no amount of developer time or effort can ever ensure it always works.

Merging WebKit and KHTML, or at least ensuring they are reasonable compatible so you can put the user agent for Konqi on WebKit or something would make Konqi a viable alternative - as it's rendering engine has a serious market (and mind) share.

Wouldn't backporting improvements made in KHTML to WebKit make sense as a first step towards integration?

by Anon (not verified)

"Wouldn't backporting improvements made in KHTML to WebKit make sense as a first step towards integration?"

This presumes that SadEagle and CareWolf have Webkit commit rights, and I'm not sure if this is the case.

If they don't, they should have - they have proven themselves *more than* worthy.

by D Kite (not verified)

I don't think it has anything to do with commit "rights".

It has more to do with willingness to contribute to Apple Corporation.

Derek

by SadEagle (not verified)

Not quite. It has to do with my lack of willingness to be an unpaid employee to Apple Corporation. There is a difference there.

by Scott (not verified)

> my lack of willingness to be an unpaid employee to Apple Corporation

Well, is KHTML paying you? My guess is that it's not (though you and other contributors to KDE are doing a wonderful job). So I would take payment out of the equation and focus on which project is most advanced and valuable to KDE users.

by SadEagle (not verified)

There is a difference between being an unpaid contributor and an unpaid employee.

by Arnomane (not verified)

I totally agree with you but I still have to admit that from a user point of view I'd love to see a true partnership and merge.

I assume you don't want to become an employed contributor to Apple but expect Apple to be more of a KDE contributor/supporter.

When I look at this page here http://ev.kde.org/supporting-members.php I see quite some names but there is one missing... ;-)

Apple got a great gift from Free software projects (not only KDE also FreeBSD, GNU...) and certainly would have been unable to code anything remote equivalent to Mac OS X (when they started coding Mac OS X, Apple still had quite some financial constraints and although I don't now any figures the developer team was probably a small one at that time).

Although they never did violate any free software licenses there's a huge difference between just acting according to the licenses and to contribute to a project. And as Apple loves to say that they support Open Source well they should actually start doing it really.

The most easisest thing for Apple would be becomming a supporting member of KDE eV. Much more complicated for them (mentally) would be turning Webkit also spiritually into project of equal partners (Apple, KDE, Trolltech, Nokia...).

(Although I don't know any current details of the Webkit SVN I am quite sure that some smart police of internal branching would make it possible that every company/project can provide a Webkit-product at their favourite fixed date).

by jospoortvliet (not verified)

Much more complicated for them (mentally) would be turning Webkit also spiritually into project of equal partners (Apple, KDE, Trolltech, Nokia...).

That is exactly what is happening right now. Apple is working rather closely with KDE, Trolltech and Nokia to merge things and keep it all together. Being an unpaid Apple employee simply isn't the question - Apple is paying ppl to work on OUR webbrowser tech - that's imho a better way to look at it.

by D Kite (not verified)

This is a classic misunderstanding of free software community and project.

It isn't about a product. It is process. There is no way at all to describe the process that has led to WebKit as anything but insulting and demeaning to free software developers. Everything in the way that tree has been maintained demonstrates that Apple considers it a one way flow, and that they would gladly have free contributors to their bottom line. That is their right, and our mistake here was not seeing that whole situation for what it was.

I don't want a browser. I want free software. I want a healthy community. I want a development structure that allows encourages excellence and pride. From that has come a kick-ass browser. And from that process will continue to come a kick-ass browser, among other things.

Primary focus on product harms the process.

Derek

by Scott (not verified)

This criticism isn't very constructive or specific. How about some concrete ways in which Apple can improve the Webkit community/process?

by SadEagle (not verified)

"Apple is working rather closely with KDE, Trolltech and Nokia to merge things and keep it all together"

Incorrect. Outside of JS and plain talking, Apple has never worked closely with KDE.

by SadEagle (not verified)

No, I see no reason for Apple to support KDE --- they have their own UI, their own OS, and it's their right (actually, an obligation, being a public company) --- to care for their interests. What we have to realize, however, is that they are acting in their interests, not ours. They may coincide, for short or long time, but what a lot of people are suggesting is giving up control over a portion of our destiny. That may work out OK, but we have to be honest about this.

by jospoortvliet (not verified)

I think the issue isn't that you become an unpaid employee of Apple, but that Apple becomes an unpaid contributor to KDE. Not taking advantage of all the resources they, Nokia, TT and others pour in KHTML/Webkit would be a waste, right?

by AC (not verified)

My highest respect to you, and all people who are still taking care of KHTML. Apple only cares about the Webkit, and doesn't care about KDE at all (heck, they only care about $$ actually).

Thanks for still working in KHTML, even though many people really want Webkit as KHTML replacement.

by SadEagle (not verified)

To Apple's credit, they are almost as good w/commit rights as KDE is. Well, in KDE you pretty much only have to ask for an account. There, you have to submit some patches, ask for an account and fill in some mostly reasonable paperwork so their lawyers don't get a heart attack.

by SadEagle (not verified)

That would be quite good, and your overall view is spot on (I know we can make a non-trivial number of websites suddenly work by lying on our 'appName' navigator field as Safari does --- may be we should do that for 4.0?) --- except it's very difficult to do with ~4000 commits; and it might take a lot longer to do so than it was to develop them, due to changes in source structures and inefficiency introduced by them having too much manpower. Unfortunately, even when aware of our implementations/fixes, Apple folks still tend to pretty much ignore them, except, again, in the matters of JS. There has also been plenty of talk about corporate $$$ from KDE-friendly companies on this matter, but none of them actually care about this sort of thing. What they care about is the new portable/cell-phone web browser craze.

by jospoortvliet (not verified)

Well, now Trolltech is going for this, I guess there will be WebKit for KDE anyway - and even more money and ppl being poured into it. I think taking advantage if that is simply the smartest thing to do, esp with apple becoming more open and all.

by Carewolf (not verified)

Switching to WebKit is not even an option until at least KDE 4.1, so we need to keep developing KHTML, no matter what solution is picked in the end.

by Diederik van de... (not verified)

+1 Good point indeed! :-)

by Rock (not verified)

As a matter of negotiation tactics with Apple? I mean, I feel that the real problem underlying the development is that KDE is running out of options. KDE 4.0 takes far far longer than expected and most features are not ready for KDE 4.0
It was a severe mistake to put so much effort on the KDE 4 cycle, we all see the early celebration of new projects and then find out that actually nothing was coded.

by T (not verified)

"and then find out that actually nothing was coded"

Have you considered trying running a diff between KDE 3 and KDE 4? :-)

Seriously, people have given enormous amounts of their creativity, intellect, and sweat to this transition. If you can't contribute, at least please don't come swooping in to make disparaging and ill-informed remarks about the Herculean efforts that have been freely given.

I for one am awfully excited about KDE4, and I think it's just great that a release is coming soon. Free software does (& should) work differently than proprietary software, because there are no forced upgrades: "release early, release often" is a great mantra for open source. I think it's amazing that things are already looking so good. Waiting for everything to be absolutely perfect would, without a doubt, slow the project down. Once it's out there, one can hope to attract more well-informed and constructive input (hint, hint) and perhaps even new developers to the project.

Indeed, I have only one gripe with KDE4: it's looking so "delicious" that I find myself tempted to abandon my beloved 3.5 series fairly soon. I'm sure I'd then discover some things that worked better in the old days. While I'm not a GUI coder, I might inevitably be tempted into putting some work into making KDE4 better. So, see what you're doing to me KDE4 developers? You're distracting me from the things I know I _should_ be doing, instead you're tempting me into helping out with the free desktop! Bad bad! I just hope I can hold out a bit longer! Aiiieeee! (Sudden sickening silence, then the sounds of KDevelop being fired up.)

by Diederik van de... (not verified)

> Have you considered trying running a diff between KDE 3 and KDE 4? :-)

lol..

I read somewhere that the change "from KDE 3.5 to KDE 4.0" already takes more commits then "starting from nothing to KDE 2". :-) So keep up the good work everyone!

by jospoortvliet (not verified)

Well, with a staggering 300-400 commits a day (387/day on average, last week), the amount of work going into KDE 4 sure exceeds everything KDE ever did...

by Anon (not verified)

"As a matter of negotiation tactics with Apple?"

What does this even mean?

"I mean, I feel that the real problem underlying the development is that KDE is running out of options."

Huh? Options for what?

"KDE 4.0 takes far far longer than expected"

It was scheduled for October 23rd. If it all goes to plan, it will be roughly seven weeks late. Even if it doesn't go according to schedule, it will only be a few months late. Whoop-de-doo.

"and most features are not ready for KDE 4.0"

s/most/some/

And developers have known this and have been saying this since the beginning of the year:

http://www.kdedevelopers.org/node/2600

Most of the KDE3 features weren't all magically present in KDE3.0, either, oddly enough!

"It was a severe mistake to put so much effort on the KDE 4 cycle,"

The Armchair Developer has spoken! What exactly should have been done, then? You're already whining about how KDE4.0 is taking oh-so-much-longer than scheduled, and now you want to delay it even further while continuing to use a less productive toolkit that Trolltech no longer officially supports?

"we all see the early celebration of new projects and then find out that actually nothing was coded."

Such as?

"The real question is whom to fire 'cause Raptor was not delivered. Development by obscurity does not work."

The developer of Raptor is not paid for his efforts, so he can't be fired (and firing the lead developer of a project generally just makes it later, fyi). Also, Raptor is developed in public SVN:

http://websvn.kde.org/trunk/playground/base/raptor/

by Kevin Kofler (not verified)

The problem with the missing features is that some of the features currently missing in KDE 4.0 are stuff which was there in KDE 3! (See e.g. Quanta completely missing.) Regressions are bad. Every such regression is going to be a real problem for distributions like Fedora which want to pick up KDE 4 early. And if distributions can't ship 4.0, who is going to be able to use it? Few people are able and willing to compile KDE themselves. It takes hours.

by jospoortvliet (not verified)

These apps can be run in KDE 3.5.x form, so that's not a regression and distros can just ship it. Sure, there ARE real regressions, like printing, but that'll be fixed asap.

by Kevin Kofler (not verified)

By the way, I sense a name conflict which is going to confuse KDE 4 users a lot: one of Soprano's dependencies is also called Raptor and it has nothing to do with Raptor the menu!
http://librdf.org/raptor/

by Sebastian Kügler (not verified)

Honestly, why do you spread this utter bollocks?

* KDE does *NOT* take far longer than expected, start reading the release schedule and compare it to the initial one.

* It was *NOT* a mistake to put such effort into KDE 4. Sure, it was risky to do all those changes 'at once', but it has great results, it has strengthened our community and it shows KDE has a great vision for the future.

* "... actually nothing was coded". Really, pull your head out of your arse and look at what's actually being done (hint: last week had more than 2700 commits to the codebase).

If you're only here for spreading FUD and annoying people (like you do with your other responses, then just go away. We don't need you.

by Div (not verified)

The guy has a point, I just tried the resent beta of KDE4, is much an alpha or maybe a pre-alpha quality.

/mw worrie.

by Anon (not verified)

"I just tried the resent beta of KDE4, is much an alpha or maybe a pre-alpha quality."

This I agree with, but "Rock" barely had anything even resembling a point - he was incoherent, confused and/ or factually wrong in almost every assertion he made. Don't encourage him, please :)

by Div (not verified)

I agree with him in the part where he says that to much effort have been taken to Plasma, it will be nice in the begining but desktop widgets are so 2003 now.

The whole Plasma idea as the desktop core I think it was a misstake, why spend so much time in something the user will get tire of in a couple of hours like plasma?

I really hope there is an option to turn off all the cluttered animations.

by Anon (not verified)

"I agree with him in the part where he says that to much effort have been taken to Plasma"

Are you kidding me? Plasma (and the Oxygen style) have received, by a very big margin, the *least* amount of development time, and from only a tiny core of contributors!

"it will be nice in the begining but desktop widgets are so 2003 now."

And yet, Dashboard and Compiz Fusion's screenlets are immensely popular. Heck, I've heard many people say that the sole reason they don't switch from GNOME to KDE is that KDE doesn't have arbitrarily resizable icons on the desktop.

"The whole Plasma idea as the desktop core I think it was a misstake, why spend so much time in something the user will get tire of in a couple of hours like plasma?"

Again, hardly any time has been spent on Plasma, and people tend to interact with their desktops/ panels/ taskbars/ launchers/ menus on a pretty frequent basis.

"I really hope there is an option to turn off all the cluttered animations."

Congratulations on avoiding the obvious term "bloated" and substituting an even more non-sensical pejorative in its place (how can an animation be "cluttered"?) What are you expecting Plasma to look like, anyway? Spaceships and hamburgers whirling round the screen every second of the day? And Aaron has already said that Phase will eventually allow animations to be shortened or switched off centrally.

No wonder you found "Rock"'s meanderings so compelling; you're almost as misinformed as he is!

by Div (not verified)

"Are you kidding me? Plasma (and the Oxygen style) have received, by a very big margin, the *least* amount of development time, and from only a tiny core of contributors!"

The Plasma Idea started two years ago, but it could only be implemented with QGraphicsView and that came with Qt 4.3, that makes me thing that the whole Plasma idea is just for Qt marketing and not for the sake of the users, after all, tha idea came from a Qt employee.

"And yet, Dashboard and Compiz Fusion's screenlets are immensely popular. Heck, I've heard many people say that the sole reason they don't switch from GNOME to KDE is that KDE doesn't have arbitrarily resizable icons on the desktop"

Im sure they are popular, but not in a scale to base YOUR WHOLE DESKTOP ON IT, it is ridiculous.

"Congratulations on avoiding the obvious term "bloated" and substituting an even more non-sensical pejorative in its place (how can an animation be "cluttered"?) What are you expecting Plasma to look like, anyway? Spaceships and hamburgers whirling round the screen every second of the day? And Aaron has already said that Phase will eventually allow animations to be shortened or switched off centrally."

If every widget, icon, window gets animated it gets in your way, when I use KDE3 is the firsth think I do is turn off animations, now take the small amount of animations in KDE3 and multiply it in an exponential way, you get KDE4 animations, come on, even OSX is moderated in that kind of stuff and surely they have the power animate every pixel if they want.

My point is, Plasma is a KDE4.1 material not KDE4.0, some how is being pushed to much, why? I don't know, well, I know, but I don't think is apropiate to come out with it.

by Anon (not verified)

"The Plasma Idea started two years ago, but it could only be implemented with QGraphicsView and that came with Qt 4.3, that makes me thing that the whole Plasma idea is just for Qt marketing and not for the sake of the users"

Aaron wanted Plasmoids to be rotatable and scalable - like GNOME's desktop icons, and OS X's dashboard widgets - and QGraphicsView gives this automatically. With increasing LCD resolutions, resolution-independence is becoming more and more necessary. I can't even begin to imagine why you think the entire KDE4 desktop metaphor was created just to promote Qt, which hardly needs promotion. It's just a complete non-sequitur. Frankly, I doubt many of TrollTech's customers even use QGraphicsView - it's very niche, and a large amount of Qt apps are simply simple GUI apps that just use the common-or-garden buttons, listviews and other ordinary widgets.

"after all, tha idea came from a Qt employee."

Let's name names here: You are referring to Aaron Seigo, who is sponsored by TrollTech to *work on KDE*. He is also noted for his huge contributions to kdelibs (which don't affect TrollTech in any way, shape or form), and is also *KDE*'s e.v. President. Attempting to portray him as some kind of TrollTech puppet is both insulting and, worst of all, based on some of the most tenuous "logic" I've ever seen. Come on!

"Im sure they are popular, but not in a scale to base YOUR WHOLE DESKTOP ON IT, it is ridiculous."

Why not? If you're going to base your desktop on something, why not make it flexible and powerful widgets /applets that can be re-located seamlessly where the user wants them, be it on the desktop or their panels? "Everything is a Plasmoid" is a wonderfully elegant way of approaching things. Also, "WHOLE DESKTOP" is slightly misleading choice of words, given that the desktop is such a tiny portion of the KDE project.

"If every widget, icon, window gets animated it gets in your way, when I use KDE3 is the firsth think I do is turn off animations, now take the small amount of animations in KDE3 and multiply it in an exponential way, you get KDE4 animations"

What are you basing that on? It's certainly not my experience of KDE4. Drop the silly hyperbole ("and multiply it in an exponential way!!!1"), please :) And as noted, it will eventually be switch-offable - the KDE guys are not stupid, and not given to wasting resources unnecessarily. With the increasing prevalence of laptops running on batteries and thin-clients running on networks, they'd have to be very silly to make this unconfigurable - and the KDE guys are not silly.

"My point is, Plasma is a KDE4.1 material not KDE4.0, some how is being pushed to much, why?"

I'll agree that Plasma won't be anywhere near it's full potential by 4.0 (it may well, frankly, suck outright), but I have no idea what you mean by "some how is being pushed to much, why?". Can you explain, please?

"I don't know, well, I know, but I don't think is apropiate to come out with it."

Hey, you've already accused the lead developer of perverting the KDE4 desktop into a mere marketing vehicle for TrollTech, so why be coy now? ;) Seriously, you obviously have a loss of trust in the KDE guys at the moment, and rather than keep your fears to yourself where they can fester, why not detail them and let the KDE guys explain their side?

by Div (not verified)

Good points you have there, here is my worries in a bigger scope and why I think Plasma is affecting:

Right now Computers sellers like DELL and HP are finally shipping Computers with Linux, and right now they ahre shipping mainly GNOME as the base desktop. Is it because GNOME is better than KDE? I don't think so, both are good and better than what vista have to offer, so what's the reason? I think is stability, GNOME may be using an inferior toolkit like GTK but is stable, Versión 2.20 it is really stable, now lets see the impressions we are getting from KDE4.0, one word: unstable.

If KDE4.0 is going to be unstable then Im afraid it can miss the bubble is coming for the Linux Desktop in the next year, will KDE4.1 be more stable? surely, but knowing in the history of KDE releases time line, it will take at least another 8 months after the release of 4.0 and mean while the compentence is not sleeping.

So the delays and unstability Plasma has brought are dangerous, I'll waith for the final release to see how stable can get, but I don't have much hopes right now. The features Plasma will bring are nice, but certainlty I don't think it would be that necessary for KDE4.0 sucess, because at the end it is something the users already have seen in other desktops, vista, osx, gnome, etc.

Im hoping to be wrong at the end.

by Anon (not verified)

Cool, now we're getting somewhere :)

"Right now Computers sellers like DELL and HP are finally shipping Computers with Linux, and right now they ahre shipping mainly GNOME as the base desktop. Is it because GNOME is better than KDE? I don't think so, both are good and better than what vista have to offer, so what's the reason? I think is stability, GNOME may be using an inferior toolkit like GTK but is stable, Versión 2.20 it is really stable, now lets see the impressions we are getting from KDE4.0, one word: unstable."

Ok, this is certainly true: there has been a huge upswing in GNOME adoption, and the recent DesktopLinux survey - if the results are accurate, which is not necessarily the case - shows that for the first time, there appear to be more GNOME users than KDE ones. However:

Dell (and, presumably, HP)[1] are mainly offering GNOME because it is the default desktop of Ubuntu which, regardless of whether it deserves it, is currently The Favourite distro, by a landslide. Frankly, I don't think Dell even care about what desktop they are deploying - in their IdeaStorm poll, people wanted Ubuntu, and I think Dell just went with that. GNOME is certainly more stable than KDE4.0 but of course, that's why we have the KDE3.5.x branch which is still pumping out huge amounts of bug fixes (and thus, extra stability). In a nutshell - I don't think that the stability of GNOME 2.20 as opposed to KDE 4.0 is the driving force behind the choice of GNOME (which, as seen in my footnote, hardly has a monopoly in the hardware world) for Linux laptop vendors, so I'm not seeing too much damage there.

"If KDE4.0 is going to be unstable then Im afraid it can miss the bubble is coming for the Linux Desktop in the next year, will KDE4.1 be more stable? surely, but knowing in the history of KDE releases time line, it will take at least another 8 months after the release of 4.0 and mean while the compentence is not sleeping."

This presumes two things that I believe are erroneous:

1) that this year is going to see a tidal wave of adoption of Linux by the masses ["200X will be the year of Linux on the Desktop" has been a running joke now since X = 0 ;)]; and

2) If KDE misses this wave, then this will cause massive, long-term (irreparable?) damage to the KDE project.

I dispute 1) because we've been hearing it for years and, though the signs pointing towards it coming true are stronger than ever, it still seems to be way, way off - Dell's own figures suggest that Linux sales are strong, but nowhere - nowhere - near toppling Windows.

I dispute 2) because I have seen many *new* linux users - new computer users, in fact - given a choice between GNOME and KDE, and *choose KDE*. There is a certain mindset that seems to attract people to either of GNOME and KDE, and it seems to be a human universal - put simply, a sizeable chunk of the population (not just power-users, as the stereotype would have us believe) will always prefer KDE to GNOME (and another chunk will always prefer the reverse). Thus, I don't really see an exodus from KDE to GNOME any time soon. Note that if, say, 10,000,000 people start using the Linux desktop this year and only a tiny minority - 10%, say - choose KDE, *KDE will still have grown*. I can't remember who said it - it may even have been Aaron again - but the Linux desktop is not even close to being a zero-sum game yet, and even if GNOME grows faster, *both desktops are still growing*.

Additionally, say KDE4.2 *really* delivers the goods - say it's fast, lighter than GNOME (a distinct possibility - figures show that KDE3.5.x and GNOME are very evenly matched in terms of RAM usage, and Qt4 offers an instant and automatic reduction in RAM usage), has powerful but elegant apps with consistent and well-designed UIs (we already have powerful apps; most just need a little nip-tuck here and there, and better defaults) incredibly easy to administer (KDE's configuration is a major plus here - Kiosk lets you sculpt it just the way you like it. You can even configure it to be less overwhelmingly configurable, if you want!) well-integrated and flat-out beautiful (go Oxygen Team!) - do you think that people will still avoid switching to it, just because 4.0 was rough and unstable? I'd say that anyone who does avoid it for this reason was simply looking for an excuse not to use it, and were a lost cause anyway.

"So the delays and unstability Plasma has brought are dangerous, I'll waith for the final release to see how stable can get, but I don't have much hopes right now. The features Plasma will bring are nice, but certainlty I don't think it would be that necessary for KDE4.0 sucess, because at the end it is something the users already have seen in other desktops, vista, osx, gnome, etc."

Ok, the "delays and unstability" of KDE4.0 are definitely *not* solely the result of Plasma - in fact, I'd say that it has hardly contributed a thing, and that it was the massive work of porting to Qt4 and the new KDE4 APIs that has consumed most of the time. Remember - there are over 5 million lines of code in KDE's svn, and Plasma is a only a few tens of thousands - if that. As a concrete example of things other than Plasma contributing to delays - just a week ago or so (a very short time before the originally planned release date of October 23rd, mind), a huge change to KDE occurred - the merging of the new KConfig branch. This was something that people wanted for 4.0, and had absolutely nothing to do with Plasma.

"Im hoping to be wrong at the end."

I think you are going to be right on a lot of things - KDE4.0 will be *very rough*, and will likely suffer a big PR backlash, especially from that minority of GNOME users that still think the Desktop War is going on and that KDE must die. Commercial vendors will avoid KDE4 for a long time, probably at least a year. GNOME will gain users at the expense of KDE.

What I don't agree with is that Plasma is going to be the major cause of delays and instability for 4.0 - frankly, there's a million things that will be wrong for 4.0, and the majority will be unrelated to Plasma in any way shape or form. Also, as noted, even if Plasma had been feature-complete months ago, there were a whole bunch of things that still needed to be done before 4.0 and which wouldn't have been done for October 23rd.

I also don't think that KDE4.0's roughness will cause irreparable damage - it will be a big, shiny, and humiliating black-eye, sure, but no one has ever died from a black-eye :) KDE's technical infrastructure is top-notch (part of what makes KDE apps so feature-rich - Qt4 and kdelibs are so comprehensive that they do much of the heavy-lifting for you) and it's development is healthier than ever - just look at the graph:

http://kde.mcamen.de/statistics.html

and I have no doubt that the devs can hack and chip and polish on the rough and jagged piece of glass that will be KDE4.0 until it becomes a jewel. And then, at the risk of sounding like a troll - it will be *leagues* ahead of GNOME, and people will see what they've been missing and switch to KDE in their droves ;)

In short - short term future looks bleak; long-term future looks rosy. Don't panic :)

[1] - Note that the EEE PC - which has received an order for 1Million laptops from Russia - uses KDE by default. I think the intel Classmate does, too (?)

by Kevin Kofler (not verified)

What do you call the initial schedule? KDE 4 has been planned for early 2007, and even before that for fall 2006, and I think at one point in the distant past even 2005. Compared to the first actual schedule, it isn't slipping much, but coming up with said schedule in the first place has taken ages.

by jospoortvliet (not verified)

The first real release schedule targeted October 2007. Yes, ppl have talked about a release before that, but hey, ppl always talk. The project never made a decision to release beginning of 2007 or something like that. We have a delay of 2 months on a 2 year schedule. MS has a delay of 3 years on a 2 year schedule. Who does better?

by Kevin Kofler (not verified)

But the years of completely unscheduled development, combined with the nonsense ETAs given by even some high-ranked KDE people, for example http://vizzzion.org/stuff/presentations/LWE2006-Utrecht/10YearsOfKDE.pdf which claimed "Will be available in first half of 2007" (and that was Sebastian Kügler, not a random dot.kde.org commenter), is exactly what gives people the perception of slipping.

by Iuri Fiedoruk (not verified)

Is just me that thinks using webkit is a good idea because it will share developers time and resources between KDE, Trolltech, Nokia and Apple?

by Annony (not verified)

It isn't just you, but it also isn't a good idea. None of these companies necessarily care about khtml in the long run. Certainly Apple hasn't been very respectful of khtml developers (read some of the earlier comments). Do you really want to tell people who are doing this *for fun* that they should instead go and do what some usurping company wants? Let them decide.

Respect your developers, they're doing it because they love doing it. Telling them to throw out their codebase because somebody with more money forked it...

If there's a way to merge, I'm sure it will happen. But I think we should make sure that it is also best for KDE. Apple has a desktop already, they don't need ours.

by Level 1 (not verified)

does this mean the end of kicker? Should we be saying our fond farewells?

Goodbye, old friend!