APR
30
2008

KDE 4.1 Alpha 1 Is Out

The KDE Community is happy to announce the first preview for the upcoming KDE 4.1, due in late July. KDE 4.1 is based on Qt 4.4's goodness, bringing performance improvements, WebKit, widgets-on-canvas and other goodies. Also new is Dragon Player, a KDE 4 port of the codeine video player which is famous for its simplicity and ease of use. KDE 4.1 Alpha 1 ships with Akonadi, the new data storage framework for our beloved PIM applications. KDE-PIM will also see its first KDE 4 release with 4.1, but is not yet based on Akonadi. More planned and already implemented features can be found in the KDE 4.1 Feature Plan. The Plasma desktop shell has just undergone major surgery, so expect some additional breakage there.

Comments

Just what we needed -- major surgery and additional breakage.


By ac at Wed, 2008/04/30 - 5:00am

Yep, that was needed. We did the following two things:

- Porting plasma to Qt 4.4 new widgets on canvas support. Previously, it was not possible to use Qt widgets in Plasmoids. Plasma in 4.0 needed its own layout system, and widgets. Now we can use all the gorgeous Qt and KDE widgets. Better integration, much faster development, smaller codebase, and instantly lots of new features.

- API review. This is more a mid-term thing. For the 4.2 timeframe, we would like to merge parts of Plasma into kdelibs. Also, before a wider adoption of the Plasma libs, we wanted to make sure that the API is intuitive to use and well designed. The changes necessary to make it a nice, consistent and easy to understand interface will definitely pay off long term. Aaron Seigo has just written an interesing blog about that btw.

All that nearly in time for an "Alpha1". In fact, both, the WoC porting and the API changes are in by now. That's where we can get back to stabilising. Not a surprising strategy WRT to release plans. That said, the current state of things feels very good, at least on my box, and starting in May there are a solid two months to stabilise everything even further.


By Sebastian Kügler at Wed, 2008/04/30 - 5:00am

> Yep, that was needed. We did the following two things:

Yes, sure. But from my POV, plasma has been the component which receives most criticism in KDE 4.0.x. So finishing the announcement for the first 4.1 alpha with "expect more breakage in plasma" is IMO "suboptimal".

Alex


By Alex at Wed, 2008/04/30 - 5:00am

Thats what KDE 3.5.9 is for.

KDE 4.0.3 is usable(have been using it as my only desktop for over a month now, and have not had any serious issues with it.)

Anyway, I downloaded the Suse 4.1 live cd. It had a lot of problems with plasmoids(none of them worked) but the basic desktop worked perfectly. It also looks amazing.


By DR at Wed, 2008/04/30 - 5:00am

If people using an alpha get mad because Plasma is somewhat broken, then its their own fault. Breakage while implementing new features are what alphas are all about in the software world. (Though in the past we've been spoiled by high-stability KDE pre-releases.)


By Jonathan Thomas at Wed, 2008/04/30 - 5:00am

People using alpha getting mad?
KDE 4.1 Alpha1 is released just to be a preview for 4.1
it 'is' alpha.this is 100% usual that an Alpha software includes broken stuff.this is why its called 'Alpha'.


By Emil Sedgh at Wed, 2008/04/30 - 5:00am

In case you haven't noticed, you guys agree.


By bsander at Wed, 2008/04/30 - 5:00am

> If people using an alpha get mad because Plasma is somewhat broken, then its their own fault.

I think the point being made is that Plasma was considered by many to be unfinished by KDE 4.0's release and responsible for a lot of its problems. Everybody unhappy with KDE 4.0 was told that KDE 4.1 was the "real" release, and that KDE 4.0 was just a preview.

So to find out that Plasma is undergoing yet more big changes rather than stabilising just makes a mockery of everybody who was fed the excuses about 4.0 and the promises for 4.1.

> Breakage while implementing new features are what alphas are all about in the software world.

Oh, I was saying that for 4.0. I was told I didn't understand open source and that betas and release candidates are appropriate for breakage while implementing new features too.


By Jim at Wed, 2008/04/30 - 5:00am

"So to find out that Plasma is undergoing yet more big changes rather than stabilising just makes a mockery of everybody who was fed the excuses about 4.0 and the promises for 4.1."

Dude, why don't you just wait until 4.1 is released? Basing your critique on Alpha-software is disingenuous at best.

"Oh, I was saying that for 4.0. I was told I didn't understand open source and that betas and release candidates are appropriate for breakage while implementing new features too."

4.0 was and is a major undertaking of moving the project to a whole new foundation. And 4.0 achieved that. Is it functionally equivalent to 3.x? No. But 4.1 should take big steps in turning KDE4 in to a robust everyday desktop.

And to respond to your whining... Again: Wait until 4.1 is released, before you start whining how Plasma is "broken", OK? Because as things are right now, you have exactly zero idea that is it breoken or not. And before you say something like "Well, the announcement of 4.1 Alpha says that it's broken!"... Yes, it might be somewhat broken in 4.1 Alpha. But Alpha is not the final release. Hell, we might get second Alpha, followed by few betas, followed by one or more release candidates.

In short: Wait until the software is actually released, before whining about that software, OK?


By Janne at Wed, 2008/04/30 - 5:00am

You seem to be more eager to parrot the party line than to read the comment you are replying to.

> Dude, why don't you just wait until 4.1 is released?

Because the point is not that Plasma is currently unstable, the point is that after telling everybody that 4.0 was the broken release and 4.1 was for stabilisation, they did the exact opposite and went ahead with more major changes to Plasma.

Whether they pull it together and fix things before 4.1 is released is beside the point. Porting to new APIs while adding features does not increase stability, it does the opposite. The fact that they are willing to do that shows that the claims that 4.1 was the stabilisation release was just an excuse. The truth is that 4.0 was simply unfinished and moving to 4.1 is a case of completing all the development they wished they had done for 4.0.

If the development process continues at this rate, people will be complaining that 4.1 is unstable and they'll get told that 4.2 will stabilise things.

> 4.0 was and is a major undertaking of moving the project to a whole new foundation. And 4.0 achieved that. Is it functionally equivalent to 3.x? No. But 4.1 should take big steps in turning KDE4 in to a robust everyday desktop.

Jonathan pointed out that alphas are for risky development, and I pointed out that means little because betas and release candidates are also for risky development as far as KDE is concerned.

Your soundbite about KDE 4.0 being a major undertaking really has no relevance to that point, it's like a generic press blurb you copy & pasted.

> In short: Wait until the software is actually released, before whining about that software, OK?

I'm complaining about the development process, not the software.


By Jim at Wed, 2008/04/30 - 5:00am

> I'm complaining about the development process, not the software.

Your complaint about the development process, while perhaps both entertaining for some and frustrating for others, is based on a terrible misreading of what has been said.

> Because the point is not that Plasma is currently unstable, the point is that
> after telling everybody that 4.0 was the broken release and 4.1 was for
> stabilisation, they did the exact opposite and went ahead with more major
> changes to Plasma.

Nobody said 4.1 was a `stabilization release'. They said Plasma would become what it was meant to be much more so in 4.1 than it had time to in 4.0. They also said that, since there will have been more testing, more bugs will have been found and fixed.

> If the development process continues at this rate, people will be complaining
> that 4.1 is unstable and they'll get told that 4.2 will stabilise things.

If the process continues at this rate, we'll be light-years ahead of the competition in just a few releases. The rate at which Plasma has been developed from nothing to a working, if not-as-featureful, replacement for KDesktop to what it is now is absolutely ridiculously awesome, for any company or community.

Software is continual stabilization. 3.5.9 is as stable as it is because it has had *6 years* of development on it at this point, 3 between 3.0 and 3.5. And that *wasn't* as major a re-architecting (the previous one having been KDE 2).

> Jonathan pointed out that alphas are for risky development, and I pointed out
> that means little because betas and release candidates are also for risky
> development as far as KDE is concerned.
>
> Your soundbite about KDE 4.0 being a major undertaking really has no
> relevance to that point, it's like a generic press blurb you copy & pasted.

It is a fact that a major re-architecting will introduce bugs. Amusingly, you make that same point earlier on. But now you say it is irrelevant to the fact that 4.0 was buggier than 3.5.9 that KDE 4 is a major re-architecting? Hopefully the paralleling there makes you see how silly that sounds. New platform releases are not for `risky' development, they are for total restructuring, which will, yes, introduce new bugs.

Finally, I've been using KDE4 on laptop and desktop for a few months (latter) and a few weeks (former), and while Plasma started out a little buggy, it's been doing fine for me recently. And yeah, it looks gorgeous. To the point where a Mac fanboy walked past the other day and went `wow, Linux is looking really good...'


By Antonio at Wed, 2008/04/30 - 5:00am

> Nobody said 4.1 was a `stabilization release'.

Wow, are you kidding? I heard it practically daily in the run-up to KDE 4.0 when everybody was complaining about it being unstable. It was the standard excuse for why 4.0 was so buggy.

> If the process continues at this rate, we'll be light-years ahead of the competition in just a few releases.

I had that opinion during KDE 3. When I saw the state of KDE 4 and the change in attitude that caused it, I changed my mind. I'm now using GNOME, and I have to say, KDE isn't the racehorse it used to be. I have to keep hold of KDE for work purposes, but I'm not using it on a daily basis any more.

> Software is continual stabilization. 3.5.9 is as stable as it is because it has had *6 years* of development on it at this point

Correction: 3.5.9 is as stable as it is because it has had 6 years of development *that wasn't focused on major re-architecting*.

Just the mere act of development doesn't make software more stable. It depends what *kind* of development it is. Some can make it more stable. Some can make it less so. The post-4.0.x work that has gone into Plasma so far is the latter.

> But now you say it is irrelevant to the fact that 4.0 was buggier than 3.5.9 that KDE 4 is a major re-architecting?

No, I'm saying that what he said isn't a response to what I said. He's just trying to talk over me with pro-KDE talking points without listening to what I'm saying.


By Jim at Wed, 2008/04/30 - 5:00am

It will come as a surprise to nobody that I basically agree with you. The Plasma project could be used as a case study of why hacking is not a valid software development method.

I have to disagree slightly about 3.5.9. It isn't really that stable and has serious regressions in the desktop which were not fixed because the developer was moving to Plasma. The reason that 3.5.9 is somewhat unstable is that changes were made faster than bugs were fixed and the result is instability.

Konqueror 3.5.9 crashes several times a day. Fortunately, this is Linux and it crashes cleanly and doesn't usually bring anything else down with it. KDE almost never crashes. Sometimes Konqueror crashes Kicker (which restarts). But stability is lacking in 3.5.9. I don't know the reason, and it would be hard to debug since the crash handler often reports that a trace is not available because the stack was corrupted.

IAC, what the development process needs is some design work -- avoid unpremeditated code!! Yes, I have an example: Bug 159492. Don't want to say anything to insult the person that wrote that piece of code, but I have to say that if a lot of the code needs that much work, we are in real trouble. My advice as a former volunteer TA is to decide exactly what you want your code to do *before* you start writing it. Trying stuff to see if it works is not a valid software engineering method -- although you should test your changes before you commit them. To do it this way and still collaborate, is going to require changes in our development methodology as others have already suggested.


By JRT at Wed, 2008/04/30 - 5:00am

"Plasma project could be used as a case study of why hacking is not a valid software development method."

You remind me of Mr Magoo..


By Richard Dale at Wed, 2008/04/30 - 5:00am

Umm... seems your KDE is b0rked? Or is it your distro's fault? My konqueror crashes as often as Firefox crashes / locks up (I use konqueror for 90% of my browsing). And in one Konqueror window I can open more than 20 tabs, so far no crash for few days.


By Ummm at Wed, 2008/04/30 - 5:00am

>Konqueror 3.5.9 crashes several times a day.

By any chance do you have Adobe Flash (and possibly the other versions of flash as well) installed and enabled by default? I've found that to be the cause of pretty much ALL Konqueror crashes/freezes (in the case of a freeze, you can often kill nspluginviewer, I believe was the name, and Konqueror will become responsive again). I disabled plugins globally in Konqueror after Adobe broke Flash in all non-gecko browsers on Linux and have had about *one* crash since then (Facebook's massive amount of js really slows down Konqi). KDE 3.5.9 is EXTREMELY stable for me. I haven't had kicker or anything else crash (that I can remember). I've had this session running for 7 days and 9 hours so far (suspending to ram at night, so maybe half that time doesn't count, give or take some hours) with 2-3 instances of Konqueror running without any crashes (I have 4 running atm, but this instance I just opened and will probably close after I finish reading the Dot and the blogs).

If you're having lots of crashes it could be your configuration (if the files in ~/.kde have been carried through LOTS of upgrades it's possible for things to just get brittle), how your distro compiles/installs KDE (ridiculous compiler flags, patches, things just installed oddly, etc), hardware failure (bad ram can lead to very hard to diagnose crashes... I've had ram that caused the system to crash when doing certain things making me think it was a bug in the software), or just really bad luck even.


By Kit at Wed, 2008/04/30 - 5:00am

Yet Flash is a feature most will come to expect as part of the standard browsing experience. Flash doesn't crash IE, Safari, or Firefox. If Flash is causing crashes on Konqui, and Konqui alone, then perhaps that needs to be looked into.


By T. J. Brumfield at Thu, 2008/05/01 - 5:00am

The Konqueror devs are trying to find ways to work around the respective bugs in the Flash plugin, though optimially Adobe would actually care about releasing a fixed version.


By Kevin Krammer at Thu, 2008/05/01 - 5:00am

What percentage of the browser market does Konqueror have? If it only crashes in that one browser, then I wouldn't say it is Adobe's problem to fix. How is it that other browsers can load the plugin with no problems?


By T. J. Brumfield at Thu, 2008/05/01 - 5:00am

> If it only crashes in that one browser, then I wouldn't say it is Adobe's problem to fix.

As long as the problem is in the Flash plugin code, which only Adobe has access to, it is Adobe's problem to fix.

When possible developers working on nspluginviewer obviously try to use workarounds to improve the situation until Adobe has fixed the cause of a problem, but of course the possibilty of such workarounds depends on the nature of the problem.


By Kevin Krammer at Thu, 2008/05/01 - 5:00am

> As long as the problem is in the Flash plugin code, which only Adobe has access to, it is Adobe's problem to fix.

Except the problem likely isn't in their code. They released the plugin to operate using Netscape plugins. Other browsers who follow that standard, implement the plugin just fine. If Adobe had a problem with Flash, it would be noticed by the countless users who use it every single day.

Clearly the nature of the problem is how Konqueror implements plugins.

This is one of the many reasons I will never uses Konqueror as web browser.


By T. J. Brumfield at Thu, 2008/05/01 - 5:00am

> Except the problem likely isn't in their code

judging by the issues that have already been fixed in the plugin, it is quite likely that most remaining issues are bugs in the plugin as well.

> If Adobe had a problem with Flash, it would be noticed by the countless users who use it every single day.

As we know it is noticed by countless uses who happen to not use Gecko.
E.g. one bug was that the plugin accessed GTK functions without calling gtk_init(), which only works if something else has already called that function.
Even a non GTK based Gecko engine would have failed that.

The fix was to properly call the init function, as documented in the GTK API docs.

A recent discovery is that the plugin calls a plugin host function and cannot handle user agent strings which include "like Gecko" but requires to have just "Gecko". However, any later call of that function can return the correct user agent, cleary indicating that there is a bug in the plugin code responsible for the first call.

> Clearly the nature of the problem is how Konqueror implements plugins.

Clearly the nature of the problem is how Adobe's Flasg plugin developers assume things outside the definition of the plugin API instead of properly using it an other APIs (see GTK misuse above)

> This is one of the many reasons I will never uses Konqueror as web browser.

I use Konqueror as my main web browser every single day, even with Adobe's Flash despite Adobe's best efforts to make this impossible.
It is awesome how Konqueror/KHTML/nspluginviewer developers manage to find workrounds and outside fixes for problems Adobe seems to be uncapable of fixing where they originate.


By Kevin Krammer at Thu, 2008/05/01 - 5:00am

> Adobe seems to be uncapable of fixing where they originate.

To be fair it was Macromedia who wrote it in the first place and someone just needs to work a bit with there other products (Director does provide there a fantastic impression how software should _not_ be) to see that it was there philosophy to deliver such kind of software ;)


By Sebastian Sauer at Fri, 2008/05/02 - 5:00am

Actually plenty of the mess in Flash nowadays was introduced after Adobe taking over Macromedia. It is the versions since then (7) which turned Flash into an apparently unmanageable kitchen sink of browser plugin meets media player meets multimedia toolkit framework. To make matters worse (likely related to the worsening of the code base) Adobe doesn't even bother to offer code of Flash version after 7 for commercial licensees anymore resulting in unsupported platforms being stuck with a more and more unsupported version 7 (a mainstream example is Flash in Opera for Nintendo Wii).

That and Flash being used for the majority of ads makes me wish it'd rather die today than tomorrow.


By anon at Fri, 2008/05/02 - 5:00am

Thanks for the details which I wasn't aware of. Somewhat frustrating for those who depend on it and a good example why closed / non-free software does provide more problems then to solve any :-/


By Sebastian Sauer at Sat, 2008/05/03 - 5:00am

If you want to see how many hacks needed to make nspluginviewer works in Konqueror, you can refer to Lubos Lunak's blog. http://www.kdedevelopers.org/node/3162 (Guess what? Opera is also not supported)

When something is badly engineered and closed source, you can only a hack to make it works. And this is the fault of those who implement the plugin.

Do you know that people with non-32 bit machine (e.g. AMD64) also got headache to enable flash in Firefox? (no 64-bit flash plugin so far) And they curse Adobe even more than us Konqueror users because of that.

I heard Adobe has released their swf spec, hopefully it will make Gnash/swfdec better and other browsers can utilize this to provide better flash.


By fred at Fri, 2008/05/02 - 5:00am

It doesn't work in Opera either


By Velvet Elvis at Fri, 2008/05/02 - 5:00am

> Konqueror 3.5.9 crashes several times a day

doesn't seem here with my 3.5.9 setup but that for sure doesn't mean, that they are not there.

> hard to debug since the crash handler often reports that a trace is not available

often is not always what are imho good news cause we would only need a valid backtrace one time :)

> serious regressions in the desktop

in Konqueror or in the desktop+kicker? Please don't understand that wrong (I just try to find the reason for that), but may it be the case, that your distributor did there something by there own? I ask cause according to the 3.5.8=>3.5.9 changelog ( http://www.kde.org/announcements/changelogs/changelog3_5_8to3_5_9.php ) there was no work done on desktop+kicker and therefore I doubt that there are NEW regressions (or better ask myself how there could be any).


By Sebastian Sauer at Thu, 2008/05/01 - 5:00am

in case of interest, that site leads to reproducible crashes of Konqui 3.5.9 here:

http://www.debianhelp.org/node/12647

but i'm feeling forced to state, that kde 3.5.9 including konqui runs pretty fine and stable here - not to compare with that infantile plasma-thing called kde4

sorry, of course there are a lot of users with the urgent need for continuously having a weather-forcast while working at their computers. Maybe only Meteorologists like me are not able to understand that ;-)


By Moische at Thu, 2008/05/01 - 5:00am

That link work without problem for me, which leads to two possibilities. Either the problem has been fixed post 3.5.9, and require a distribution which backport such fixes or you have to wait for a possible 3.5.10 release. The second possibility are that your distribution simply are broken.


By Morty at Fri, 2008/05/02 - 5:00am

You are an annoying person who is not only looking at a gift horse's mouth, but downright insulting the donor!

You are also an ignorant person who thinks re-architecturing and cleaning up the code should never be done. You believe incremental updates lead anywhere in terms of features irrespective of the starting codebase.

They don't, the limit was 3.5.9, which is brilliant and extremely stable, except where the architecture found its limits -- PIM, the desktop/panel.

The goal is to have plasma _in kdelibs_ which means binary compatible, which means that the API must be _right_, because there won't be changes in the next five or so years. This is the _correct_ way of doing it. And for such a thing as the widgets, which must cater to morons also (sorry, Aaron had some social-network-name for them) this is _hard_. And this is what will make all the alternatives irrelevant in the medium term.

So if they want rearchitecturing, let them. Be happy that they do, and that they don't listen to the uninformed opinion of armchair commentators such as you. I am using the svn snapshot, and although plasma is pretty broken just right now, by the next snapshot it will be fixed. And going back to KDE3 is painful: KDE4 apps are, by now, downright better.

I will however grant one thing, much trolling came from the way the 4.0 release was handled. From a developmental point of view, the KDE team did the right thing, from a please-the-trolls point of view, no so much. I suppose if you care about what you do, sometimes, you need to do the right thing anyway...


By hmmm at Wed, 2008/04/30 - 5:00am

> You are also an ignorant person who thinks re-architecturing and cleaning up the code should never be done.

Please don't make stuff up. I don't think this at all and there's nothing in my comment to suggest it.

> The goal is to have plasma _in kdelibs_ which means binary compatible, which means that the API must be _right_

I thought the party line was that KDE 4.0 was for application developers to port to KDE 4 rather than end-users? More incompatible changes in the API? Will KDE 4.1 be for application developers to port rather than end-users as well?

Don't you understand that fixing the API at this point only strengthens my claim that KDE 4.1 was just an excuse for KDE 4.0 developers to finish their work?

> And for such a thing as the widgets, which must cater to morons also (sorry, Aaron had some social-network-name for them) this is _hard_.

Which is why hacking on widget code shouldn't happen in the main branch, on the code responsible for vital UI features like the taskbar, when you are supposed to be stabilising things!

> And this is what will make all the alternatives irrelevant in the medium term.

I'm sorry but there are a *lot* of widget systems about, Plasma really isn't that compelling, and unlike KDE, the alternatives don't let work on the widget systems drag them down.

The core idea of Plamsa is fine, I have no complaint there. But the way development has occurred, and the way KDE has relied upon it prematurely, has turned it into a complete albatross around KDE's neck.

> although plasma is pretty broken just right now, by the next snapshot it will be fixed.

People have been saying this ever since the first alphas of KDE 4.0. "Don't worry, the beta will be stable". "Don't worry, the next beta will be stable". "Don't worry, the release candidate will be stable". "Oh shit, we're not going to get this done for 4.0... I mean, KDE 4.0 isn't for end-users, dummy, you should wait for KDE 4.1".

> I will however grant one thing, much trolling came from the way the 4.0 release was handled.

A troll is somebody who posts insincerely in order to annoy people. Very little trolling came from the KDE 4.0 release handling, although false accusations were abundant. Why can't you understand that people have legitimate complaints about this?


By Jim at Thu, 2008/05/01 - 5:00am

> Why can't you understand that people have legitimate complaints about this?

Legitimate based upon what? What do you believe gives you the right to continually complain on this site about the way KDE is developed? Furthermore, what do you hope to achieve by making complaints here?


By Paul Eggleton at Thu, 2008/05/01 - 5:00am

See, I understand that it might disturb you as a concept, but ignorance is not a point of view.

- The rewrite of plasma was planned from the start to happen upon the release of Qt 4.4.

- Aaron, being the maintainer of kicker and co. said it was a nightmare. And he is a pretty smart guy. If he says: "this must be replaced by something new and sensible", odds are, he is right. Also, there are _now_ way more devs on plasma than there ever were on the old shell. Can you guess why? Oh, no, you can't because in your vision of the "right" universe, we would have no more devs and a half-working shell and a depressed and frustrated Aaron. IRL, you are a PHB, right?

- It is not about having a widget system, it is about having a general, all-encompassing widget system which builds upon all the other widget systems. And the design of which also allows, as a side effect, the rebuilding of a desktop shell. And you know what, they are pretty much there.

- No complaints are legitimate here, none, never. Unless you are a dev. You might state that you are unhappy, you might voice opinion, but to complain, you have no rights.

- This is free software, this is not Apple inc. where all the mishaps and difficulties of development are hidden from the customer's view, you get to see them, people debate in the open about problems and difficulties. This is how you get better software in the end. Thinking of free software as a product gives you wrong ideas. It has no products, just projects that at some points become better than products.


By hmmm at Fri, 2008/05/02 - 5:00am

"Wow, are you kidding? I heard it practically daily in the run-up to KDE 4.0 when everybody was complaining about it being unstable. It was the standard excuse for why 4.0 was so buggy."

The thing I have heard is that 4.1 is the version that should match 3.x in terms of feature-set. And now that they add new features, you whine? And if they didn't add new functionality, you would whine that "we were promised that 4.1 will have all these new bells and whistles!".

And like I said: are there any indications that 4.1 will be buggy? Yes, 4.1 alpha has some issues. That's why it's an alpha-release, and not the final gold master. What you are doing is complaining because the next major version of KDE has new code and functionality, and some bugs in the alpha-release. YOu somehow extrapolate from the alpha-release that "4.1 will be buggy!", even though several people have said that Plasma in SVN-snapshots is coming along nicely.

And what is this about the developement-process? Are you somehow shocked that there's major rewrites and like between major releases?


By Janne at Wed, 2008/04/30 - 5:00am

> YOu somehow extrapolate from the alpha-release that "4.1 will be buggy!"

I repeat myself:

> the point is not that Plasma is currently unstable, the point is that after telling everybody that 4.0 was the broken release and 4.1 was for stabilisation, they did the exact opposite and went ahead with more major changes to Plasma.

> Whether they pull it together and fix things before 4.1 is released is beside the point. Porting to new APIs while adding features does not increase stability, it does the opposite. The fact that they are willing to do that shows that the claims that 4.1 was the stabilisation release was just an excuse. The truth is that 4.0 was simply unfinished and moving to 4.1 is a case of completing all the development they wished they had done for 4.0.

> Are you somehow shocked that there's major rewrites and like between major releases?

I'm shocked that some people are saying that they are stabilising Plasma while simultaneously making major changes to it.


By Jim at Thu, 2008/05/01 - 5:00am

"I'm shocked that some people are saying that they are stabilising Plasma while simultaneously making major changes to it."

So, do you believe that the plasma (or Kwin, or any other part of KDE) that we had in 4.0 will be the version we will have for all eternity? Like I said, major version-number (like 4.1 is) are there for major new features and functionality. And I bet that Plasma in 4.1 will be more stable and functional than Plasma in 4.0 was. They can do both: make big changes and make it stable.

Like I said before: instead of whining about Plasma, how about waiting until 4.1 is released before passing judgment on it?


By Janne at Fri, 2008/05/02 - 5:00am

Even though reality seems to have no impact on your opinions, try to spin that:

This rewrite was planned from the start to happen upon the release of Qt 4.4. The "breakage" was fixed within a week of the rewrite.

The rewrite was mandated, obvious, necessary because of WoC, and the fact that much of the logic originally in plasma is now found in Qt. The fact that this results in significantly simpler plasma code does not affect your opinion that obviously it _must_ introduce new bugs and hinder stability.


By hmmm at Fri, 2008/05/02 - 5:00am

I can't help but get the impression that once a piece of software reaches the point where you can get work done reliably and securely, you wish it to freeze in place and undertake no more improvements what-so-ever. Is this correct?

As to your complaints about false promises, I remember much whining about missing features and Kickoff being the default menu, but nothing about stability. The promises made in response were that 4.1 will reach feature parity with 3.5, and perhaps even surpass it in some areas. It seems to me this may be bothering you, since if Plasma does surpass KDesktop and suffers no major stability issues as consequence, you will no longer be able to make complaints that will not be regarded as issues of personal taste.


By yman at Wed, 2008/04/30 - 5:00am

> I can't help but get the impression that once a piece of software reaches the point where you can get work done reliably and securely, you wish it to freeze in place and undertake no more improvements what-so-ever. Is this correct?

No.

> As to your complaints about false promises, I remember much whining about missing features and Kickoff being the default menu, but nothing about stability.

In that case, by all means browse each and every release notice that appeared here on the dot, on Reddit, on Digg, on Slashdot... there were a hell of a lot of complaints, you'd have to be blind to miss them.


By Jim at Thu, 2008/05/01 - 5:00am

>> I heard it practically daily in the run-up to KDE 4.0 when everybody was complaining about it being unstable. It was the standard excuse for why 4.0 was so buggy.

Without a source your statements are useless. Also, if 4.1 is more stable than 4.0, then saying that 4.1 will be a stabilizing release is accurate. Without a change in Plasma API, perhaps it could have been more stable. But if we were after ultimate stability, KDE 4 would never have been started, so this argument is stupid.

>> I'm now using GNOME

And yet you still feel the need to troll the KDE boards. Apparently you aren't happy with Gnome either, otherwise you wouldn't ever look back.

>> I have to keep hold of KDE for work purposes, but I'm not using it on a daily basis any more.

The fact that you apparently need KDE for work is telling.

>> Correction: 3.5.9 is as stable as it is because it has had 6 years of development *that wasn't focused on major re-architecting*.

No shit. That's the point.

>> The post-4.0.x work that has gone into Plasma so far is the latter.

If you're not involved, you can't possibly know that. I'm using a snapshot of 4.1 right now that is much more stable than 4.0.x ever was.


By Leo S at Wed, 2008/04/30 - 5:00am

> > I'm now using GNOME

> And yet you still feel the need to troll the KDE boards. Apparently you aren't happy with Gnome either, otherwise you wouldn't ever look back.

Sorry, you're the one using the classic trolling tactic of only partially quoting the other person. The quote in full:

> I'm now using GNOME, and I have to say, KDE isn't the racehorse it used to be. I have to keep hold of KDE for work purposes, but I'm not using it on a daily basis any more.

It's quite obvious that using KDE is a necessity to me regardless of what I think of it or GNOME.

If you cannot debate sincerely, please just be quiet.


By Jim at Thu, 2008/05/01 - 5:00am

>> Sorry, you're the one using the classic trolling tactic of only partially quoting the other person. The quote in full:

The funny thing is I included the rest of your quote 2 lines down from that line.. But I guess you didn't read the rest of my post...


By Leo S at Fri, 2008/05/02 - 5:00am

>> Nobody said 4.1 was a `stabilization release'.
>
> Wow, are you kidding? I heard it practically
> daily in the run-up to KDE 4.0 when
> everybody was complaining about it being
> unstable. It was the standard excuse for
> why 4.0 was so buggy.

I'm with Jim here... You guys can deny it all you want, but this was the impression regular users where given. ("Yeah, KDE4.0 wasn't so good, but you just wait for KDE 4.1, man! I'll be great!")

KDE4 is shaping up to be a huge disappointment, and somewhat of a disaster. Oxygen looked good in theory, but the implementation is really bad with buttons and icons all over the place, the active window-problem, and different styles everywhere -- and plasma is still unstable. Then you go and introduce a new API? WTF?!

Is there *ever* a time to polish the way KDE4 look? Honestly, KDE4 looks like shit now. The GNOME-folks OTOH have done a great job polishing their DE, making it look really attractive and coherent (albeit somewhat of a throwback to Win2K looks-wise). It's time for me to upgrade, and as bad as KDE4 is looking now, I'm moving to GNOME...


By blah at Thu, 2008/05/01 - 5:00am

Just stop with the FUD. Before the API refactoring, Plasma was quite stable. And now, just slightly a week after said API refactoring, Plasma is stable once again. The API changes were necessary, and once all the obvious bugs that were introduced in the applets due to the API change are fixed, the base is only much more stable because it is now relying more on Qt4 code and less on custom Plasma code built to duplicate Qt4 features.

Read this for more, and keep in mind that what you are throwing a shit fit about is alpha software: http://aseigo.blogspot.com/2008/04/on-recent-libplasma-changes.html

"Disaster"? Hardly. Currently alpha-staged software? Yes. There is no doubt in my mind that KDE 4.1 will be awesome.

P.S. The default Plasma theme will go under the polishing board, and Oxygen issues such as the inactive window "problem" are being addressed.


By Jonathan Thomas at Thu, 2008/05/01 - 5:00am

WTF? Major complaint about KDE 4.0 is because it needs more features once we had in KDE 3.5.9, especially desktop and kicker. And now they want to make the desktop better and you still complaining?

And about KDE4 looks like shit, did you live under the rock when GNOME released their 2.0, or were you with your proprietary system? GNOME 2.0 looked like the worst single GUI on earth, and yes, they didn't have clearlooks or ubuntu-human that make GNOME looks so beautiful nowadays. Give them a time dude! I guess you even never tried KDE 4.0 or KDE 4.1 alpha.

My last advice: go troll somewhere else!!! Sometimes we just hate idiot users who can only complain and whine. And also, who the heck cares if you're moving to XYZ Desktop Environment?


By fred at Thu, 2008/05/01 - 5:00am

> KDE4 is shaping up to be a huge disappointment, and somewhat of a disaster. Oxygen looked good in theory, but the implementation is really bad with buttons and icons all over the place, the active window-problem, and different styles everywhere

The "active window" problem is being worked on. Also, the appearance of Oxygen (as with any other style) is very subjective. If you do not like it, change the style.

> and plasma is still unstable. Then you go and introduce a new API? WTF?!

The last time Plasma crashed for me was during the KDE 4.0 beta phase. If you have crashes which are not caused by own incompetence (i.e. which are reproducible what I assume) report them.


By Stefan Majewsky at Sat, 2008/05/03 - 5:00am

> Wow, are you kidding? I heard it practically daily in the run-up to KDE 4.0
> when everybody was complaining about it being unstable. It was the standard
> excuse for why 4.0 was so buggy.

As others have pointed out, I would appreciate your telling me where you saw that. People were promising that 4.1 would be _better_, yes. People were saying it would have fewer bugs, because it would have been more tested, as I pointed out in my own post, something which you utterly disregarded. There is a massive difference between `yes, there will be fewer bugs' and `our focus will be stabilization rather than new features'. The former is normal for major releases, the latter is not. Stabilization is for bug-fix releases. The 4.0.3s of the world, which have shone in that regard.

> Correction: 3.5.9 is as stable as it is because it has had 6 years of
> development *that wasn't focused on major re-architecting*.

And it had the luxury of not focusing on major re-architecting because it built on the major re-architecting in KDE 2, which introduced features (KParts and KIO, for example) that other environments (Windows and such included) *still* don't have.

> Just the mere act of development doesn't make software more stable. It
> depends what *kind* of development it is. Some can make it more stable. Some
> can make it less so.

Which... Was exactly what I said. Thank you for reiterating.

> The post-4.0.x work that has gone into Plasma so far is the latter.

Well, for one thing it wasn't post-4.0.x, it was parallel-to-4.0.x. Which is important because bug fixes were made along the way. Yes, this new API has obviously re-broken some things, but the point, as, again, has been pointed out elsewhere, is that the API is hopefully going to go into long-term maintenance mode, and, by affiliation, the underlying stuff will not be *able* to be revised for five years or so.


By Antonio at Thu, 2008/05/01 - 5:00am

> it [KDE 3.5.9] had the luxury of not focusing on major re-architecting because it built on the major re-architecting in KDE 2, which introduced features (KParts and KIO, for example) that other environments (Windows and such included) *still* don't have.

Perhaps the point is that we do not have many developers (esp. when compared to M$ and Apple) and thus have to implement things in a very effective way. Many key components of KDE (such as the mentioned KParts and KIO libs) and the underlying Qt libraries allow applications to implement features with less code (and therefore less possible error causes).

Some examples from my personal impressions: To open a website in the user's favorite browser, the code is one line (KRun::runUrl()). Downloading a file from an arbitrary location (from local files to web services to SSH resources) costs three lines of code (KIO::NetAccess::download() to temporary file + error check + cleanup temporary files). Under Windows, access to files at arbitrary locations requires you to essentially write a new API for your program which links against several libraries.


By Stefan Majewsky at Sat, 2008/05/03 - 5:00am

> I'm complaining about the development process, not the software.

How would you have preferred it?

Waiting three years until a perfect KDE 4.1 is released and all Plasma issues are fixed?
Wait with the porting of Plasma to Qt 4.4 and WoC sweetness?
Break the API between 4.1 and 4.2 instead, or don't review it at all?
Label this release "KDE 4.1 this-is-really-REALLY-pre-pre- -pre-unstable-not-for-users-KRASH-alpha-release" ?

KDE 4.1 isn't going to be perfect, nor absolutely stable, but it's one step in the progress of KDE4.


By Hans Chen at Wed, 2008/04/30 - 5:00am

Pages