OSNews

Nobody will be using KDE 4 for anything that matters for at least another twelve to eighteen months.

I am referring here to large school, government or corporate deployments. So even if they are still pushing the envelope technologically with KDE4, they have plenty of time to stabilize.

For the foreseeable future, kde 3.5.8 and greater will remain the desktop platform for serious deployments.

That may change eventually and I for one cannot wait to get my hands on the exciting KDE4 platform, but it will take a while to mature. Which is perfectly fine in my book.

Cut that time frame in half and you strengthen your views.

Thats fine so why don't they refer to the release as an alpha? Its not really their strategy that bugs me its their terminology..

The phrasing is probably not optimal. I basically means that the last portions of API cleanup had happened, not that totally new concepts were introduced.

A bit later the press release says "As the KDE libraries were in freeze for big additions since the first Beta, work was mostly focussing on fine-tuning and fixing everything."

Is it just me or does "finalizing the design of the libraries" sound very much like alpha type work rather than something that should belong in beta.

Whatever it means, at least all the KDE4 betas which I have tested so far, including Beta 2 which is part of Ubuntu Gutsy, felt very much like alpha quality software. They were not just buggy and unstable as hell but essential functionality was simply missing or totally broken.

However exciting the new feature may be, I don't think that premature status changes are helpful to their advance. If they really ship in January, they will sure ship an unfinished product.

I've been hearing that for quite a while now, and it's entirely unlike my experience with (self-compiled) KDE 4. Sure, it's not rockstable, and some day something works, another day it doesn't. And plasma of course has been horribly unstable. But most other things work pretty well. Most apps in KDEbase, kdegraphics, kdeedu, kdegames - actually, almost every app has been usable and rather stable for quite some time yet. So there must be problems with the Kubuntu packages, or you only tried the few things which aren't ready yet - like plasma.

No no no no.

Finalizing the design happens in the end, just before releasing the stable product.

You start by getting an idea, then you elaborate it, then you construct it and then you put it out in real world. These phases can either be hard phases (e.g. the Waterfall model which is a disaster) or they can be cyclic. The latter one is true for all variations of Agile Modeling (including AUP). Most of the design happens during elaboration ((pre-)alpha) but the finalizing happens during transition (the phase just before releasing the stable product).

It's just you.



It's about both.



It's nice to know that OSS and closed-source software apparently has a common denominator.

It's not 'these days', it is just like it has always been. Release early and release often.

For sure, but whats the problem releasing under the alias alpha instead of pretending you have beta quality?

we HAVE beta quality, it is only plasma which isn't up to Beta right now. Almost all other apps are very usable. But because plasma is so visible, everyone who tried KDE 4 had a bad experience because they did nothing more than click around 2 minutes on plasma, concluded it sucked and they went to flame KDE on OSNews.

KDE and GTK use the same font rendering engine, which is provided by the FreeType libraries, not the toolkit. You can enable subpixel rendering and all that other stuff, but you may have to tweak your ~/.fonts.conf

Weird... Do you really thing that GTK, which uses exactly the same font rendering library as Qt renders fonts differently? There may be different default freetype settings, like having kerning on or off, but
that's all. Every toolkit on X11 uses freetype. And no font current technology uses concepts from the antigrain website article.

I really know that parts of Xft/RENDER/Pango/GTK/freetype do render fonts, sure more than everybogy here, lookung my scorepoints down to 0, but obviously you do not. You assume there is some secret organizations force QT and GTK to use only that path? Thats the point, time to mature. 20 years old assumptions that every glyph is a just squarish box with integer aligned corners is wrong. Freetype guys do good work, but all other (GTK/Pango/Xft/RENDER) do not take patches that output gamma correct blending, subpixel accurate placement( kerning). And I hope that KDE devs do something with it.

With all your knowhow on Xft/RENDER/Pango/GTK/freetype etc. you sound just like the right man for the task. We're all behind you.

Boudewijn doesn't asume anything, he knows - he's a rather prominent KDE developer. And KDE uses what is available - and freetype is the only real option on linux. If you go and develop a better alternative, chances are KDE will start to use it. Until then - freetype it is.

Don't expect KDE to start working on a font rendering infrastructure - that's by far not KDE's job. We don't develop drivers for the kernel or X either, we develop higher-level gui libraries and a desktop+applications. Trolltech, the Qt developers ARE involved in font rendering, a lot - the current Pango etc stuff is for a big part developed by them. If you want to improve upon that, go and contact the font ppl who work on that kind of stuff.

I am not sure but I think you are mixing this up with Harfbuzz, the text layout engine, which is based on large Trolltech contributions.

Hmmm. I think it is looking beautiful. I agree that the 3.5x is tiring. But, this new stuff looks fresh and wonderful.

The digits in the clock are struck though because it is a flip digit clock design.

like this

http://www.zillion.co.nz/listing/4090719/viewphoto/283051/

Also, IIRC, they plan to replace the default clock with a much simpler design at some point, so that it will be much clearer on a smaller panel.

The digits in the clock are struck though because it is a flip digit clock design.

Yes, but what's wrong with ordinary non flipping digits. There are no technical reasons for doing this on a computer screen, and this flip design is not common enough in real life to motivate doing it this way to help people identify it as a clock. It just makes it harder to read for people that is visually impaired.

It is also a matter of style, these flip clocks belonged to the 1970:s and 80:s, the rest of KDE 4 doesn't.

The question is should there really be a digital clock, analog clocks are much easier to read as they not only give you the actual time but also gives you a much better view of the time as an ongoing process.

I agree that the flipping digits design looks weird, like a bug almost. I contacted the artist and asked him to update it by not having a straight line gutting through all the digits but have it just so slightly tilted at each cut which in my mind would make it look a bit nicer, but he did not show much interested.

The problem with bringing real-world concepts and physical constraints into a computer GUI is that the have to be exaggerated and modified to fit the constraints and opportunities of the GUI. Flipping digits you say? Ok, so do they actually change by an animated flip? If no, drop them. If yes, reconsider as that can be highly disturbing.

Yes, I'm also looking you, you spinning desktop cube!

the flip thing is just a funny 'hey I can do this' thing, it will probably be replaced for 4.0 - or 4.1, then.

It is mainly a technology demonstrator, i.e. showing how to use SVG theming and animated state transitions with a very simple data source.

And it also serves as a nice show item.

An "all-day-long" type of clock will most certainly be visualised less intrusive.

So the default clock in 3.5 was LCD, very 80's, now for 4.0 we'll have flip digit, 70's. Can't wait for KDE 5.0 lol

Note: for god sake get rid of those huge windows'borders. Vista is a plague. Since it's out, we got at least one per day at kde-look.org and even gnome-org for almost any existing window manager.

I'm pretty sure the clock looks like that because the plasma developers wanted to test out their animation framework, and it was about the only available plasmoid they had to test with at the time. I doubt it will be the default look, although they might keep it as an option.

The widgets to me actually look really nice. Pleasent to look at, pretty clear contrast now I'd say. I like the slight hints of color there too, not overdone like the crazy blue borders in 3.5 which aren't bad but very very bright which I dunno, it just gets tiring as a default pretty quick.

My only real complaint is the tiny window buttons.

Can I still slam my mouse in to the corner? Or will I have to carefully move my mouse over the button.

I like slamming my mouse to the corner... easier...

It's not a problem, it's a theme. If you don't like it then change it

Kicker is dead, and replaced by a plasmoid in plasma which is still brand new. And probably will be rewritten before 4.1 comes out, from what I've heard.

From SVN:
> Aaron J. Seigo committed changes in /trunk/KDE/kdebase/workspace:
> Now the kicker lay down to sleep,
> I pray svn its code to keep;
> And if it crashes ere I wake,
> I pray a backtrace you do take.
>
> <fadeout>"s-i-i-lent night ... h-o-o-ly night ... doo doo doo, doo doo doooo.."</fadeout>

Who said anything about rewriting the DE? I said they would be rewriting a single plasmoid. Plasma was designed from the ground up to make replacements like that possible, since a big issue in KDE3 was that it took huge amounts of work to even make a tiny change to Kicker.

Edit: Ok, looking back at my post perhaps it does look i was saying they would rewrite plasma as a whole. That's not what i meant, sorry.

Edited 2007-10-18 05:32

nah, they won't rewrite plasma, they're just gonna improve on the current panel implementation. Sure, it's worse than Kicker - no doubt about that. But porting kicker only to remove it later on (it wouldn't have been feasible to keep it around for the whole 4.x series anyway) was considered a huge waste of time. Thus it was decided to settle with a bit less functionality for 4.0, to be fixed in 4.1.

It's temporary "programmer art":

http://pinheiro-kde.blogspot.com/2007/10/its-your-turn-to-oxygen.ht...

Pinheiro and Ruphy were discussing changing it yesterday over IRC, but I went to bed before finding out what they were planning, and when it would be done.

yes yes yes, it's horrible. I think it's a Kubuntu problem, OR you expect too much. Most applications work great here, it's just plasma which is rather incomplete, and a few other minor things.

Don't know about the number of contributors (although easy and inviting tasks like writing pet plasmoids has been responsible for inducting a few people, and David Faure adds a bunch of new SVN accounts quite frequently), but it's certainly increased the activity:

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

Draw a vertical line down from the "3.4.1" release on the graph - this is, roughly, KDE4.0's "frist post!1".

http://cia.vc/ is a nice way of tracking commits in real-time, or you can go to #kde-commits on irc.freenode.net.

Why the hell would they rewrite it? They've got all the makings of a great platform, the actual desktop itself just needs a bit of polishing. 4.0 won't be perfect, but then neither was 3.0. I'd rather have 4.0 on my desktop in January and 4.1 eight or so months later than see the release delayed far into the new year. Even if it is a bit rough around the edges, 4.0 will still be a vast improvement over the aging 3.x series.

"why should they rewrite it?..for starters it resembles Windows vista to much"

This can be fixed by theming, not re-writing. Apply make-up, rather than decapitating someone and re-growing them a new head

" plus i think its better to have a stable release than an unstable one"

A KDE re-written from scratch would introduce so much instability that it would be insane.

" i think KDE4 needs to or will need to mature before Distributions put it on there desktops, why rush a release? "

Yes, KDE4 will need to mature before distros will default to it - that's simply a no-brainer. They want KDE4 release because the simple fact of the matter is that, perversely, only a tiny minority of people bother to test Beta releases, and KDE4.0 really needs a "trial-by-fire" to hammer it into shape. Release early, and often is an old open source adage, and the KDE guys have not been adhering to it as well as they would have liked. Plus, a growing number simply want it out the door - sitting on software this long without releasing builds up an unpleasant kind of psychological pressure that can lead to loss of motivation, and releasing provides - well, a release. Feature-freezes can be unpleasant for much the same reason - a feeling of slogging and slogging away, and not accomplishing anything visible.

So do you want a rewrite or a stable release? The two are mutually exclusive.



Why would they?

KDE 3 was unstable? Not from my recollection...

And why is looking like Vista a bad thing? If that's all you can criticse KDE for, it's looking good.

Dave

/me thinks that people seem to forget easily the train wreck that GNOME 2.0 was back in the day and quickly point fingers to everybody else's dot zero release these days...

Having said that, I too think that Vista is a half-baked attempt that was rushed into the market given Microsoft financial investments on it, though. However, neither KDE nor even GNOME can take advantage of that sort of funding and/or resources and so the comparison between them and Vista is moot, IMHO.

No.



It will be ready when it's ready. Deal with it.

Stacks is patented.

Stacks is patented.

Tough. Apple really do piss me off, because it's always been possible to do variations on a theme. After all, it's basically a vertical list.

As much as I admire KDE, it still looks dull, ugly and uninspired. They really need to find some graphic designers to help them give it some professional zest..........It just looks....so....boring.

Wow. Well, it does look completely different to the old KDE, it only looks like Vista in terms of the black status bar and it will only look completely different as KDE 4 development progresses.

Since you're obviously overflowing with ideas, give us a description, some artwork and some idea of what you think you'd do.

It's as horrible as it looks, and no, luckily for us all it's not final. Plasma (part of which is the taskbar) still is going through a lot of work.

I agree. And I hope the taskbar is changed before the release of 4.0. In screenshots where a lot of windows are open, it looks very cluttered. And without clear dividers between each 'button' in the taskbar, I think it would difficult to quickly find the one you are looking for.

the taskbar is really still open for major changes, as it's part of Plasma.

Hey, I added quite a few screenshots to the article ;-)