OSNews

Interestingly, when I wrote my two editorials on the future of GNOME and KDE, I already predicted that the KDE guys would never be able to release KDE 4.0 in 2007. I was almost crucified for saying it - by people here on OSNews, but also by people within the KDE community.

http://www4.osnews.com/story/16783/Has_the_Desktop_Linux_Bubble_Bur...
http://www4.osnews.com/story/16802/On_Favouritism_Apologies_and_Bla...

And look where we are now. It's not that I mind KDE 4.0 got delayed - seriously, I prefer delays over a crappy product. What bothers me is that people went all mental when I predicted these delays.

Anyway, Aaron, I know you read OSNews, so how about this [1] ? It's getting interesting now .

[1] http://www4.osnews.com/thread?249710

Edited 2007-08-31 20:16 UTC

i know, it's going to be a photo finish! =)

the tagging date has been tentatively moved back two weeks to dec 6, though, so we have a proper chance of it getting out in december.

whee!

whee!

LOL he's such a happy chap sometimes ;-)

hehe

Interestingly, when I wrote my two editorials on the future of GNOME and KDE, I already predicted that the KDE guys would never be able to release KDE 4.0 in 2007. I was almost crucified for saying it - by people here on OSNews, but also by people within the KDE community.
...
And look where we are now.

As I recall, you were predicting it would be 4Q 2008, with a decent chance of slipping into 2009, which is quite a bit later than the delayed Dec. 2007 launch. Still, your date is probably close to when 4.1 will come out and everything will be finished up, as it sounds like 4.0 is going to be without quite a bit of stuff.

KDE 4.0 will not be released until summer 2008.

And we should care about your predictions because...?


Me, I say that the two months delay is enought to get KDE 4.0.0 out. Making random guesses is free anyway!

Edited 2007-08-31 20:29

"""

"""

You shouldn't. It's mainly important to me. That's why I kept it short. ;-)

KDE 4.0 will not be released until summer 2008.

That's a losing bet. A better prediction would be whether or not a KDE4 release makes it into the spring 2008 distributions such as Kubuntu Hardy Heron or Fedora 9 by default.

> or not a KDE4 release makes it into the
> spring 2008 distributions such as
> Kubuntu Hardy Heron or Fedora 9 by default.

Kubuntu HH is a long term support release, so KDE 4 is not planned to be the default desktop on the grounds that it is not expected to be sufficiently mature. Kubuntu HH users will of course be able to get KDE 4 applications I'm sure.

"""
That's a losing bet.
"""

We'll see. It's just a feeling that I have. We can revisit the issue next June or whenever they actually release.

And you base that on what? A hunch? Well, then, wanna bet?

It's pretty unlikely we won't have a release until summer 2008, with the release event planning and all - that makes it pretty hard to delay beyond the new date. Actually, we can't and won't.

This is the only thread that anyone will post two for the next week. KDE4.0 is on every ones mind

Perhaps there merger of WebKit for Konqueror and more is actually going to be completed for KDE 4?

Do you think that they wanted to wait for KDE 5?
KDE 4.0.0 won't get new features – just more time to finish targeted features.

WebKit will come in KDE 4.1 or 4.2.

I agree... they have the whole world watching ... it's to big a thing to mess up. Take 4

Just because there's a lot of things under the hood that aren't immediately visible to the typical user from merely looking at the GUI doesn't mean that those under the hood things won't make a notable difference in how well that visible things work for the end user.

It'd be far better to delay some reasonable amount of time and get everything anchored down and working as intended than to have to issue constant errata, and/or keep on breaking applications that depend on those unseen bits to not change behavior over time, as that directly/indirectly affects the user-experience and how they judge using the system. Also, if you make something too iffy for developers to deal with and trust, that backfires horribly as well.

If the consensus amongst the developers is that it isn't ready, I'm going on the basis that they know what they're talking about, and if nothing else, I could see it as a bit of personal/professional pride to make the judgment "no software before its ready and sufficiently stable" and applaud them for that. Plus, on a professional level, once you release something that's bad enough, it's almost impossible to earn redemption in the marketplace, so it's just bad business to release something that you know isn't ready.

Certainly, and that's why I don't expect to KDE 4 to be stable and feature complete for the next 10 or 12 months. Patience is a virtue.

Why not?

It is not about the face, it is about the soul!!

Maybe KDE 4.0 will have the same UI than KDE 3.5.*; but it does not matter when the developers have been replacing the KDE's soul to be far better than the "soul" of previous versions:

A lot of new frameworks and infrastructure are being built as the KDE 4 foundation; these changes will be the base for a lot of new applications which will bring the user with a wonderful experience.

Maybe there are not applications that use the new stuff, but with the release of KDE 4.0; all the means to jump far far from its current status will be already there.

Good job KDE guys! My hat off to you.

Edited 2007-09-01 00:03

Mac OS X, just as Windows Vista, have ton of resources and many millions on cash to avoid delays. KDE has not. Delays in KDE are comprehensible.

Well, it's not as bad as any of the delays Microsoft has given us. A two months delay is what we in Denmark call "planned delay" or "expected delay".

If KDE4 becomes as delayed as Vista then we can talk about poor management. A 2 months delay is pretty well done for a project this size. Better than 4 years late

yeah I agree. A project that took this long doesn't surprise me at all they decided to delay a couple of months. I mean they did plan the announcement for January in the first place. So it sounded like someone was planning for a delay in the first place.

How delayed was Vista, exactly? They promised a 2 year release and they hit a 2.5 year release if you consider that work was started in early 2005. You don't remember some other OSes released in that timeframe, like 2003 SP1/XP x64 (with a lot of work to support a new architecture)? You don't remember XP SP2, which was a pretty extensive modification that came out in that time frame?

Let's leave this thread to die because it's far more important to consider the good work done by the KDE devs who seem to be defining a clean and extensible architecture for the future of the linux desktop. Marketing and release-planning foibles of Microsoft are simply not pertinent to this discussion.

How delayed was Vista, exactly? They promised a 2 year release and they hit a 2.5 year release if you consider that work was started in early 2005.

Dude. We had technology previews of Longhorn in 2003, and Microsoft was talking about the thing for four years.

However, I'd prefer to leave this to die because that's what I'd call hype.

I have never taken a position to defend Microsoft, but don't compare an entire Operating System to a Desktop Environment.

When KDE has to also build a Kernel then you have a gripe.

How many official release dates did they announce? What were they?

"Don't tell me u are comparing a window manager to the whole operating system.... "

Because, you know, KDE is just a window manager. It is in no way an entire framework for a desktop environment. No. That's just what they want you to beleive. kwin, that's all there is.

"way to go dude, u kick some major ass"

Dude, where's my car?

"don't they have mental institutes for guys like u in Denmark???? "

Institutes? Don't you mean institutions? Duuuuuude.

It kind of worries me that up until recently it was expected they could meet the October deadline given how much work still had to be done and how big of a project it is. That's poor project management.

I take it you're new to open source development then? It's little a thing called iterative improvement. It's Microsoft and Apple that have big-bang, hard and fast ship dates on their wallcharts - and they missed them by miles.

So when FOSS does it, it's iterative improvement, but when a corporation does it, it's missing their ship dates by miles.

K, got it.

So when FOSS does it, it's iterative improvement, but when a corporation does it, it's missing their ship dates by miles.

Their development models are like night and day. Do you see every single internal alpha, beta and nightly build at Microsoft and Apple? No, because they're private. KDE and other open source projects do their development in public, with public point releases.

K, got it.

------------> Concept










------------> You

You do see OS X alphas if you're in the inner list and are a large developer who has paid to see them. You won't see the nightly builds, but you'll definitely see monthly or sooner.

No they are not. Their software licensing models are, but I doubt their actual development processes are really all that different.
Microsoft's and Apple's development models differ only from FOSS in that development happens behind closed doors - behind the scenes, people are frantically writing code, testing code and brainstorming ideas in much the same way. Microsoft actually did release quite a number of public beta versions of Vista, and criticising MS for missing a shipping date but not KDE using the same criteria seems a little biased to me.

In either case, shipping a product that isn't ready is worse than shipping a finished one a bit late. I have no problem with either MS or KDE or anyone else taking a bit more time to make sure the product is OK before releasing it. End users should not be treated as beta testers on shipped versions (one reason why it could be argued that MS should have waited a little it longer still on Vista).

Microsoft never ships within 2 months of promised delivery. Microsoft kept pushing the release date of Vista until few months before release. And at that time Vista was four years late.

KDE4 is not four years late. It looks like it will be two months late which is way better than anything Microsoft so far has managed to do. They are never less than 6 months late, and often years late.

Never ships within 2 months.. never? Strong words.

This isn't about Vista though. This is about KDE. Just because Microsoft has some management layer issues, that doesn't make it OK for KDE because "at least they're not Microsoft".

You show me a size of this magnitude, complexity, diversity that has a perfect schedule.

A revolution of sorts. A complete rewrite. A diverse worldwide group of developers from all over the globe. Many working for free in their spare time.

You're simply asking far too much. Period. Vista was delayed how much? Granted, this is not an OS it's a DE...But there's a fraction of the resources allocated to this project vs Vista.

So...yeah...quit being so judgemental.

Excuse me for being worried that they didn't see this coming earlier or refused to.

Hold on. If KDE also developed Qt and the entire frameworks that it entails then you have an even greater footing to stand upon.

Novell and other corporations are helping KDE succeed. This isn't just a bunch of bachelor to ph.d computer scientists writing a desktop environment.

They aren't writing and develop Xorg or the Linux Kernel or the Qt Frameworks, etc.

They are augmenting Qt with their own Frameworks, leveraging Xorg and any Kernel that currently it runs atop and much more.

There definitely are quite a few developers contributing to KDE.

There are definitely quite a few developers contributing to Gnome.

From the Dot KDE Digest latest:

3095 by 244 developers, 7218 lines modified, 1544 new files.

Hell: If you want to know how fractional NeXT Software Inc., had in developers I'd be glad to tell you.

Let's just say they had no more than 50 for an entire Operating System, Development Suites (Openstep Tools and WebObjects combined) to accomplish Openstep and KDE has 244 developers just for the Desktop Environment.

Either NeXT had nothing but the worlds finest developers with an incredible ability for design and implement or KDE has a shitload more resources.

Don't get me wrong: KDE is excellent, but definitely not Revolutionary. Shit even OS X has devolved to satisfy the Mac faithful in quite a few areas. It's taken 10 years to shed that waste to finally start seeing a direction that one can be excited about.

Too bad a lot of talented Cocoa developers left Apple during the transition because they couldn't stomach waiting a decade for the transition to be completed when we were told it would be a few years.

In short: Promise less and deliver more.

The difference is, the Next developers were full-time, and almost all KDE developers do their work in the evenings, after their day job. That makes an enormous difference.

The difference is, the Next developers were full-time, and almost all KDE developers do their work in the evenings, after their day job. That makes an enormous difference.

Both projects show that small numbers of people with talent can do amazing things. And that with all the money and people in the world, Microsoft can still deliver a real, expensive late dog of a desktop environment with Vista.

A more interesting comparison might be how Apple compares with NeXT, now that many times more engineers work on Mac OS X than ever worked on OpenStep. For instance, I don't find the Apple Finder/Dock is more usable than the NeXT workspace manager (I think it's worse), and all the extra Apple people don't seem to have done much innovation in that area.

I like KDE4's Dolphin with column view browsing, split views, integration with the semantic desktop etc. Maybe that will end up being more usable and powerful than the Mac OS X Finder.

Granted, this is not an OS it's a DE...But there's a fraction of the resources allocated to this project vs Vista.

Why the hell are you still comparing it to Vista then? Hell, why is anybody on this thread?

KDE has enough money in the coffers to keep it going (from Trolltech, obviously), so does GNOME. Apple/MS's problems are worlds apart from GNOME, KDE, and every other desktop environment. Apple has to make and sell a whole OS, Microsoft has to make and sell a whole OS.

Meanwhile, KDE and GNOME, simple desktop environments, just have to be coded by their respective projects; other companies (Novell, Red Hat) have to sell the OSes that use these desktop environments.

So does MS have an excuse for having delayed Vista? Hell yes, they're managing a whole OS. Does Apple have an excuse for having delayed Leopard? Different reasons, but they're perfectly legit (they're managing an entire computing system).

Does KDE have an excuse for having delayed KDE4? Yes, but far, far less in quantity than Microsoft or Apple.

So just stop with the comparisons between Vista/Leopard and KDE/GNOME, all of you. All comparisons are illegitimate and useless to the discussion at hand.

Apple and MS only code for a single platform i.e. their own, so they have control of their destiny
KDE/Gnome are multi-platform.

A couple of months delay is not a problem for any system. MS and Apple get into flames because they commercially hype their systems to try and stop people moving to the competitor. Its not surprised they get hoisted on their own petard and quite right too

MS and Apple get into flames because they commercially hype their systems to try and stop people moving to the competitor.

Apple doesn't have a competitor, since they sell an entire desktop computing system (hardware+software, not either-or). Their closest competitors are Dell+Windows, HP+Windows, Lenovo+Windows, etc., not Dell by itself nor Windows by itself.

Microsoft's competitors are Novell, Red Hat, Xandros, Mandriva, Canonical, Linspire, etc., and their commercial offerings of Linux-based systems. Being companies that are all out to sell operating systems to the public, isn't it their job to hype their own products against the competitor anyway?

And what does the commercial hype of either company have to do with this topic anyway?

Edited 2007-09-01 21:29

1. afaik, no.

2. I think there will be easy video players for 4.0, but I'm not sure. Maybe someone will write some after the release, as I guess that would relatively easy with Phonon and all...

I'm not a KDE fan either, but that's an easy comment to answer. The KDE community doesn't speak with one voice. Some want KDE to be released soon, even if it's unfinished (so that it can possibly get into distros, especially long term support distros) while others don't want this until it's ready (it would divert attention from the 3.x branch which works for them and may give KDE bad press if the 4.x is a big yawn or too spartan).

KDE 4 would never make into any real LTS release. It would need extensive testing for quite some time. I want KDe to succeed, because if ut succeeds then that means gnome has to bring its A game to compete. Competition is great. There is alredy talk among developers to revamp gtk and I think that for the most part this is due to the recent developments in the qt and KDE scene. The gtk devs got toolkit envy.

Sadly, I don't think GNOME devs can compete with the likes of KDE...they both seem to have different views...For instance, KDE aims to be extremely customizable with an abundance of options..GNOME not so much. They like to keep it simple. In either case, it's a matter of choice on behalf of the end user.

Just remember though, increasing the number of options, features and choice increases the complexity, development time and likelihood of shipping with major bugs, and choices for end users should be balanced against the needs of the majority of users. It is better to split a project up into a number of projects target at specific groups of users than to create one gigantic project that tries to be everything to everyone and eventually collapses under its own weight.

Keeping it simple is a good principle - I think a lot of features have been added to KDE without any real consideration of the need for that feature to be there, the percentage of users who would take advantage of it, and how they would be affected if it wasn't there.

Gnome has taken a considered look at their broad userbase, and followed sensible design principles, whereas KDE sets out to be the Swiss Army knife of DEs, and risks being a jack of all trades and a master of none.

Choice usually involves a trade-off - if you choose KDE, you are trading off simplicity for control, which means you are trading off reliability and stability because complex software is inherently harder to test and develop than simple software.

Aside from aesthetics and usability, I've always found Gnome to be more stable and less buggy than KDE.

But if KDE floats your boat, go for it. This is just my opinion after all.

How can you be so sure of this when Kubuntu 8.04's development hasn't even started yet?

Edited 2007-09-01 00:16

Oops. I meant Kubuntu 7.10.
My mistake.

The delay isn't that big. Suse 10.3 wouldn't have had it anyway, Fedora 8 will have a RC instead of a final (with updates every week or something) and Ubuntu 8.04 wouldn't have had it anyway either.

I know the are/is thing, but fewer/less - that's more difficult for me. hey, I'm dutch, sorry!