MAY
28
2007

KDE Commit-Digest for 27th May 2007

In this week's KDE Commit-Digest: Continued work in Plasma, particularly in the clock visualisations. Kalzium uses the GetHotNewStuff framework to download new molecules for its 3d viewer, plus speed optimisations for the rendering of these molecules. The start of fullscreen support in the Gwenview image viewer. Work begins on a WebKit-based KPart. A KControl module is created to allow for easy manipulation of KWin "Composite" settings. Continued work on the OpenChange Akonadi resource which enables interoperability with Exchange servers. Statistics plugin for a graphical representation of connection speeds in KTorrent. Improved handling of HDR imagery in Krita. Branch created for the integration of Solid-based connection management and notification in Konqueror.

Comments

OK, you can call me a Troll, give me minus karma points, say that I'm a idiot, but... just a clock?? What's next, desktop eyes? :)

Don't take me wrong, plasmoids are nice and beatifull, but I was expecting changed in the core interface of KDE. Ff you look at the plasma page it says that plasma is about rethinking the desktop paradigm, and I belive that is waht people are expecting from it.


By Iuri Fiedoruk at Mon, 2007/05/28 - 5:00am

oh, come on. they're DEVELOPING the damn thing. i donät care what you say or what your intentions are, i think you still are acting like a troll.


By Viktor at Mon, 2007/05/28 - 5:00am

As I said, yes, you can call me troll, I know I'm kind of trolling.
I just tought, and see that that guilt for that is plasma homepage(!) [that's the troll part, I admit, I took what was written there as true], that the work could be on UI bases for KDE, not plasmoids. Plasmoids can be done throught Karamba also, so I TOUGHT it was low priority for Plasma.

Obviously I was wrong :)


By Iuri Fiedoruk at Mon, 2007/05/28 - 5:00am

Think of it like this: Sometimes before you can take a step forward, you must take a step backward.


By Sutoka at Mon, 2007/05/28 - 5:00am

Well, you've totally misinterpreted the nature of plasma then. Plasma is a replacement for kdesktop and kicker. It is designed to replace the DESKTOP metaphor, not the whole UI. KWin is still there. The applications will behave in much the same manner you'd expect applications to behave. What will change is the taskbar, the tray, the clock, desktop drawing, etc. and it's being designed in such a way to make it trivial to extend, and add bling to these components...

OS X is famous because of it's dock. It doesn't really change the way the applications themselves work from OS 9, but it changes the general feel of the system. Think of plasma in the same way.

Additionally, there are new widget drawing styles, and window decorations in the works to go with the Oxygen icon artwork that will give KDE 4 applications a fresh visual appeal - this however is in no way part of the plasma interface.

Cheers


By Troy Unrau at Mon, 2007/05/28 - 5:00am

Like it said in the video this was the first plasmoid made. Since Plasma is still under heavy development, and KDE 4.0's release is looming closer and closer, Plasma's priorities are to provide equivalent functionality that Kicker/KDesktop provided for the KDE 3 series (AFAIK). Remember that KDE 4.0s release doesn't mean that KDE 4 is done, just that it has only begun.


By Sutoka at Mon, 2007/05/28 - 5:00am

Yep, I agree, absolutely!
Actually I'm saying since long ago that Plasma (as defined on it's page) will not be part of KDE 4.0.

I just surprised that work started on plasmoids and not on KDE UI.
Aaron Seigo even showed some mockups very nice of icons with text and other things, and I belived that would be first in Plasma.

BUT, maybe the clock is just a first step or a test, and not a indicative of things to come (x-eyes, climate applet, cpu temperature..). In this case I will be happy if you guys call me a clown, a troll or whatever :)


By Iuri Fiedoruk at Mon, 2007/05/28 - 5:00am

> BUT, maybe the clock is just a first step or a test, and not a indicative of things to come (x-eyes, climate applet, cpu
> temperature..). In this case I will be happy if you guys call me a clown, a troll or whatever :)

That's EXACTLY what all the other replies AND the article in the digest said. The Clock is the first Plasmoid and there will be more to come.

> Aaron Seigo even showed some mockups very nice of icons with text and other things

If you're talking about the mouse-over icon selection menu thing, then that was discussed for Dolphin, not Plasma desktop.

It makes no sense that you're complaining about us not working on "KDE UI" and then you tell us to not work on the desktop/panel etc.


By Matt Williams at Mon, 2007/05/28 - 5:00am

Plasmoids are not just fancy widgets. IIRC kde4's kmenu and panel replacements are also plasmoids. ie why plasmoids are important.
Check this out,
http://techbase.kde.org/Projects/Plasma/Menu


By kryptik at Mon, 2007/05/28 - 5:00am

Well, troll if you must, but KDE is still months away from releasing. For the longest time, after kdesktop was killed, we didn't even have a background rendering.

Open source is not done behind the scenes, so since it is public, you get to see some of the early development screenshots. Sometimes this is a mixed blessing as you users expect something that is fully polished when the first screenshot is released. The alternative is to get no news at all, and in fact, with trolling, and user comments such as this one, the developers and writers (such as myself) become less inclined to show off early work. It's kind of sad actually.

So please, when you look at the screenshots and the videos, please look at it as a proof of concept. This is the very first visible evidence of work being done on plasma, which needs to be treated as such. There is real code powering those objects, and the groundwork has been laid for pretty much every other kind of applet you can think of, including the existing superkaramba applets. While a clock is not that exciting, it shows that much of the groundwork now exists and functions.

Additionally, kicker will shortly be put to rest, and will be reimplemented as a plasma applet. This means that while you are used to kicker being a separate application, it will now be as easily configurable, and extendable as any desktop applet ever was. The problem is, that at the moment, people are working on the backends - the stuff you don't see. For example, the panel replacement would need a number of data sources programmed - it needs stuff to feed to the taskbar, the tray; it'll need a new menu system and the menu editor will probably need to be reworked.

Sometimes you make our jobs totally not worth it.


By Troy Unrau at Mon, 2007/05/28 - 5:00am

Please don't become disheartened and withhold news / screenshots etc! For every user who doesn't get it there are lots, like me, who are waiting with baited breath for the next week's commit-digest. Sad, I know, but the stuff that's going into kde4 has got me really excited. I reckon even for many of the ungraceful comments, it's the case that the people that make them are really psyched about the work you guys are doing.


By Adrian Baugh at Mon, 2007/05/28 - 5:00am

OK, I'll call you a troll, but this comment isn't meant to bash you; please don't take anything personally.

First we complain that Plasma hasn't made any progress. When Plasma finally is mentioned in the digests, we complain that it's "just" a clock - where's the revolution?

Basically I think people (including me) has very high expectations of KDE4 and Plasma, seeing "just" a clock is disappointing. But please think of it this way: it's still under heavy development, the most work isn't even visible to the eye. The devs are giving you a chance to follow the development and give constructive criticism - if you know you sound like a troll in your post, can't you at least try to sound nicer.

At least I appreciate when people are showing screenshots and videos. Even if I don't feel "WOW" I can see what's going on and point out what I like/dislike and give suggestions. And believe it or not, but the devs are listening. Love you guys. :)

OT: This time it was actually a "wow". Too bad I get errors when I run cmakekde when compiling kdelibs from SVN (Kubuntu, following the guide on techbase). I want to try to drag and resize the clock myself.


By Lans at Mon, 2007/05/28 - 5:00am

Sure, no offense :)

I really think Plasma is doing OK, and that just a clock is OK if you plan in the long road.
My main complain with Plasma is actually their home-page.. man, I wish they never had publiced that! There are a lot of promisses there that can't be filled in the short term, and it generated too much antecipation for something that, let's take real, probally will be really functional around KDE 4.3.
Taking that aside, I don't even bother keeping the award winning and cute KDE3 interface (kicker, kdesktop, konqueror, etc), I like it a lot and my computer isn't very fast for fancy graphics (just a old Duron 1.6).


By Iuri Fiedoruk at Mon, 2007/05/28 - 5:00am

> There are a lot of promisses there that can't be filled in the short term

yes, i suppose the best thing to do is not have any goals. the next best thing to not having any goals is not to tell others about them so that nobody can help you work towards them.

you, sir, are brilliant.

> probally will be really functional around KDE 4.3

and you base that prediction on what?

> I like it a lot and my computer isn't very fast for fancy graphics

it'll work just fine on your system.


By Aaron J. Seigo at Mon, 2007/05/28 - 5:00am

>> There are a lot of promisses there that can't be filled in the short term

>yes, i suppose the best thing to do is not have any goals. the next best thing to not having any goals is not to tell others about them so that nobody can help you work towards them.

>you, sir, are brilliant.

Although I generally agree with you, you totally missed the point here. Having goals is good. Publishing goals is good. But it really depends on how you publish them. I won't argue with you about when all those things will be ready, noone could tell better than you. But from what I've heard it is pretty obvious they won't all be ready for KDE 4.0 and that is what should have been stated clearly on the Plasma page since the very first day of its existence. Telling afterwards (yes, I do follow your blog) that KDE 4 is not KDE 4.0 seems like a lame excuse. I'm not saying it is, actually I'm pretty sure that this was clear to you from the very beginning. But for most people it was not and it's only natural they are a little disappointed now - despite all the cool things there already are, they just expected more. When Apple announces features for OS X 10.5, you expect them for 10.5.0, not for 10.5.x .

All in all a communication problem and you as a developer should really not have to worry about this. Although I can understand this is driving you nuts sometimes, just try to ignore it. You know what caused it and you can avoid it the next time. There's no point in wasting time answering such posts.

Btw thank you for all the great work you did on KDE and your interesting blog posts!


By wassi at Tue, 2007/05/29 - 5:00am

Oh, and to add some really useful feedback: Please make it possible to delete Plasma widgets from the screen just by dragging them into the waste bin - would be really intuitive (not sure if OS X has it)


By wassi at Tue, 2007/05/29 - 5:00am

"Please make it possible to delete Plasma widgets from the screen just by dragging them into the waste bin - would be really intuitive"
+1 :)


By Dolphin-fanatic :) at Tue, 2007/05/29 - 5:00am

and +1 again !


By Cissou at Tue, 2007/05/29 - 5:00am

easy, keep them working. a clock is actualy GREAT, because it means that the basics work. it's like a hello world that works, it means that you can do other great stuff.

just wait and see and dont complain


By Beat Wolf at Tue, 2007/05/29 - 5:00am

Where exactly have I been complaining ?

And I know what this clock really means (framework-wise) since I know programming a little too. But it really does not matter what you show off, as long as there's not everything included that was promised on the Plasma page, people will always complain - so there's absolutely no point in listening to them. That was the essence of my post...


By wassi at Tue, 2007/05/29 - 5:00am

kde4 isn't out yet, so no one can tell if plasma will meet the expectations that plasma.kde.org is posing.
And looking at that website, i am confident that plasma will meet most of them.
Just don't expect to see hundreds of plasmoids on the release day of kde 4.0


By whatever noticed at Tue, 2007/05/29 - 5:00am

Try the kdesvn-build script instead. I used it last night for the first time on my ubuntu fiesty box and it worked without any problems.


By Chuck Shaw at Mon, 2007/05/28 - 5:00am

There's something like that? Damn, I think you just ruined my sleep tonight. ;)
I'll try that, thanks!


By Lans at Tue, 2007/05/29 - 5:00am

> but... just a clock??

the idea of plasma is to provide new ways to build things for the desktop and new ways to arrange your desktop. that is the core idea.

to create that we need a few example widgets and a few dataengines. then we exercise the framework with those and get the underlying structure in place so as to support all kinds of user visible goodness.

i know that from a user's perspective a clock isn't exciting. i'm not trying to excite you with it, so there you go, we're all even. what i'm trying to do, along with the rest of the plasma team, is make something that can produce things that will excite you. please understand that plasma is not a marketing exercise, but a technology project which means that technical goals and processes are what we aim for. exciting you will happen at appropriate points in development, but i'm not going to try and excite you with useless hacks and half-designed rubbish.

and yes, we have lots of plans as to what we will produce, but we have to get there first.

> but I was expecting changed in the core interface of KDE.

if you could understand the technical underpinnings of what we're working on and had the ability to see big picture items, maybe you'd begin to see how that's where we're going. i'm not renigging on what i've said in the past, though i'm getting the idea that people have a really hard time understanding what "getting there" looks like.

interestingly, people find the path there most uninteresting. which is ok, most good ideas are, in retrospect, not only obvious but built up from small logical steps that seem in the process to be straightforward but that end up somewhere completely unexpected. so, like a good joke, wait for the punchline. don't interupt half way through and say "it's not funny yet."

> Ff you look at the plasma page it says that plasma is about rethinking the
> desktop paradigm, and I belive that is waht people are expecting from it.

tell you what, leave the thinking to me and just sit back and enjoy the ride. it'll be a lot less frustrating for both of us. people are trying to figure out what we're doing and evidently failing spectacularly at it. that's ok, but let's roll with it and instead of getting everyone flumoxed, let's just agree to let the thinking happen on this side of the fence. in return, you'll get some nice free software. deal?

i'm sorry if that sounds harsh, but i'm tired of dealing with this.

you know, i remember in elementary and high school teachers would always pair me up for group projects with people who couldn't keep up. so i'd end up doing all the work, so as not to endanger my grade, and just tell my partner(s) to look busy but stay out of the way. i guess some things never change.


By Aaron J. Seigo at Mon, 2007/05/28 - 5:00am

"i know that from a user's perspective a clock isn't exciting. "
well, annma posted a kde4 screenshot on the planet recently, and it damn well seemed exciting to me. ;)
...later that day I discovered that I can't access kde svn from behind this evil proxy. maybe after class today I'll have to hack around that. I wanna see kde4 again!


By Chani at Tue, 2007/05/29 - 5:00am

> so, like a good joke, wait for the punchline. don't interupt half way
> through and say "it's not funny yet."

hehe thats a nice one :-)


By b at Tue, 2007/05/29 - 5:00am

Agreed. For anyone who is not a programmer, it's impossible to appreciate the degree of triumph a compiler-writer feels in getting "hello, world" to compile and execute. If that were something I had just done, the next thing I'd do is call in my next-door officemate and show it off. "Months of work for that???" would be the standard reaction from anyone who doesn't know programming.

Even for people who do program, it's pretty hard to appreciate just what's going on without familiarity with the technical details. KDE4 is clearly going to make it far easier to produce beautiful, functional applets & applications, but the closer you are to the details the easier it is to see the progress that's being made and how it's going to impact things.

So, for every 100 users who make unhelpful, naive comments (and remember they're watching in part because they are excited about what's going on), I hope you do attract the attention of one person who actually helps with some of the heavy lifting. If that happens, then I hope you'll view the publicity as a success.


By Ted at Tue, 2007/05/29 - 5:00am

I actually was looking for things less visual.

In my (humble) opinion, after reading the plasma page, I tought it was going to be like a UI plan, to be followed by applications, think of it was a set of guidelines and tools for developers to use on their GUI.
In my mind, things like a integrated find (a le firefox) would be part of Plasma, but maybe I just had misread the Plasma homepage text.

That said, yes, a clock looks nice, but I kind was expecting "Development of KDE4 has just begun, and it is during these major release cycles that we have the opportunity to retool and rethink our applications and environment at the fundamental level." :)


By Iuri Fiedoruk at Tue, 2007/05/29 - 5:00am

Yea you did misread. Plasma is the kicker/kdesktop replacement. It is going to do more then the current kicker/kdesktop (scriptable, desktop widgets), but its not something that will be used from within other apps.


By Ian Monroe at Tue, 2007/05/29 - 5:00am

well it still will have DBus interface, etc., so it still can be used from within other apps, but for notifications, etc.


By Jordan K at Wed, 2007/05/30 - 5:00am

My desktop paradigm needs a clock :)


By Cyrille Berger at Mon, 2007/05/28 - 5:00am

Neither am I a developer nor do I know how to program. I just use KDE and it's applications and find them to be so polished that I recently convinced several people to make the switch from Windows to KDE. One of the programs we really like is Basket. I was so glad to see this great piece of work Sébastien Laoût created and just hope that some developer steps up and fills the gap. Please read the Commit-Digest to find out about all the great features that were planned for Basket 2.0! Hopefully someone takes heart and continues the development of this great application.


By kde-user at Mon, 2007/05/28 - 5:00am

I agree - I find this program very useful, and use it regularly when planning my Road to KDE 4 article topics. I can only dream of what could be done with this program with a qgraphicsview...

If only I was better at coding, I'd take up this program...


By Troy Unrau at Mon, 2007/05/28 - 5:00am

Troy,

While I find Basket very useful, your work on "The road to KDE4" is at least as valuable to the community, and I for one would hate to see you stop publishing those.

On the other hand, the best way (in my humble opinion) to become stronger at coding is to jump in at the deep end and learn as you go! :-)

- Nikolaj


By Nikolaj Hald Nielsen at Mon, 2007/05/28 - 5:00am

Thanks, I appreciate the kind words :)

I have taught myself fairly well to code with a number of technologies, from databases to complex numerical analysis, however I have never been good at the UI stuff. I'll leave that to the people that can work those things out, and I'll stick to my role within KDE :)


By Troy Unrau at Mon, 2007/05/28 - 5:00am

Rather than QGraphicsView, I would use the Flake library of KOffice. I did discuss it with basket's author, and it would really fit, as throught the shapes, it would give access to texts, images, vectors, etc... with editing possibilities for each.


By Cyrille Berger at Mon, 2007/05/28 - 5:00am

Why must Basket coded in C++? With Kross and the kde-bindings to good scripting-languages (like PyKDE) it would be no problem to create a useful and more dynamic application. Also, it would be possible to outsource the development to the community, by using an extendable framework, like firefox did it. And in the end, it would be less work. So, why C++?


By Chaoswind at Mon, 2007/05/28 - 5:00am

The problem with you're argument is that you're suggesting that firefox is written in a scripting language -- it most definitely is not. It is written in a real language, compiled to machine code, and the extensions are scripted. This is pretty much the way everything in KDE works as well. See Amarok for a good example of using scripting languages for extensions within KDE.

Basket is coded in C++ because it's the core application.


By Troy Unrau at Mon, 2007/05/28 - 5:00am

> The problem with you're argument is that you're suggesting
> that firefox is written in a scripting language

Well, no. Im just say, that a application can gain advantage from outsourcing code to plugins.


By Chaoswind at Mon, 2007/05/28 - 5:00am

> The problem with you're argument is that you're suggesting that
> firefox is written in a scripting language

The platform that power Firefox (XULRunner, Gecko, GTK+/Win32/...) are written in C and C++, but Firefox, the interface, which also includes a lot of logic, are written in XUL and JavaScript. Its perfectly reasonable for a program like Basket to be written in a high-level language. (I am not necessarly advocating that, just saying; I prefer C++ myself and I bet that's the case of Basket's author.)


By fast penguin at Mon, 2007/05/28 - 5:00am

Afaik the UI in KDE is also, like firefox', mostly xml-like stuff. I'm not a developer, so I'm not sure about this, but I don't think it's that different...

Anyway, the idea from Cyrille sounds very interesting - Basket and flakes...


By superstoned at Tue, 2007/05/29 - 5:00am

no, there is a big difference. qt ui-files are xml, but they are mostly used to generate c++ code (though qt can also load them at runtime, that's not needed for most applications).

also, using only xml or a scripting language (like javascript with xul/firefox) is still not the same. xml (when used for what is intended ;)) will only increase the startup time - with a slow scripting language (or rather a slow interpreter) your whole gui may become unresponsive.


By ac at Tue, 2007/05/29 - 5:00am

So that people can use it without bogging their computers down... I need my desktop to be FAST, and I do NOT allow any program written in Java, much less any scripting language like Python or Ruby. They're fine for simple scripts, and as scripting languages to extend applications, but if someone makes a full useful app in Python, I will start looking for a C/C++ coded alternative. No apps written in scripting languages on my (ancient) PC, thank you. BTW, that's exactly what I hate about firefox. It's big, slow and bloated. I wish someone coded some efficient, open-source, multiplatform alternative browser (yeah, I know I'm getting greedy :)), so I could get rid of that piece of bloat from my desktop.


By zonk at Mon, 2007/05/28 - 5:00am

> They're fine for simple scripts, and as scripting languages to extend
> applications, but if someone makes a full useful app in Python, I will start
> looking for a C/C++ coded alternative.

Thats your personal problem. Python is more than fast enough for desktop-applications like basket. I use dozens of self-written PyKDE-Apps, without any performance-problems. Ok, of course they are slower than Native-Qt, but that doesn't count in real live.

> BTW, that's exactly what I hate about firefox. It's big, slow and bloated.

Firefox is'nt really slow, it has only a horribly thread-management, and a stupid Default-Configuration. But, yes, there are some bottlenecks, because of to much used Layers in the design.


By Chaoswind at Mon, 2007/05/28 - 5:00am

Well, if you're lucky, Konqueror for KDE 4 could be packaged up in such a fashion... it runs native on OS X and Windows now... the downside is that there are a number of libs that would have to be shipped with it, such as Qt, kdelibs, etc. which could drive up the download size. On the flipside, this libs are shared between a number of other newly portable KDE applications, such as Amarok - so that downside could be easily spun to be an upside...

Of course, we'd prefer to pull people onto the more free platforms and away from OS X and Windows... then they can have KDE in all of it's beauty.


By Troy Unrau at Mon, 2007/05/28 - 5:00am

I'm always intrigued by this apparent trade-off that computer languages must either be "expressive and rapid to develop, but slow on execution" (like Python or Ruby), or "slow and hard to develop, but fast on execution" (like C/C++). Come on, you can have *both*!

Why don't you guys try some of the functional languages out there, like Ocaml, Clean, or Haskell? You get the best of both worlds: highly expressive languages which make development a breeze, and at the same time their runtime performance is comparable to C/C++ (check the computer language shootout for numbers).

I know there was a project to add Qt/KDE bindings to Ocaml, but it seems dead. It's a pity, because I am sure that KDE could benefit greatly in supporting the functional paradigm...


By dario at Mon, 2007/05/28 - 5:00am

I know it works with KDE after a fashion - it's used in parts of Kalzium for KDE 4..., Ocaml, I mean.


By Troy Unrau at Mon, 2007/05/28 - 5:00am

I agree with you!

People say that Python or such languages are "fast enough". This is true to a certain point. When Basket implements cross-baskets searches or other advanced methods, you need resources! A good developer should not think about the application he is going to have in three months, but about the application he is aiming at!

Somebody says you can implement the slowest parts in C or C++ and plug the code into Python. Who is going to do that? You had better be motivated, especially if your objects are complex!

Python is very good and I use it everyday. But I can get much further in C++, even if it is harder at the beginning.


By Hobbes at Mon, 2007/05/28 - 5:00am

There is a wrong assomption that a an application written in python/ruby will b e entirely written in python/ruby, the database backend (hence the search engine) or a mp3 decoder or an image processing algorithm should be _written_ in a language which is translated to assembly (that can be python with gcc ;) even if I didn't try), but then the UI just need to do the glue. A lot of things are easier for the programmer, memory management, comprehensive errors, etc... And in the end, it doesn't result in a slow program, but it definitively results in something that was easier to debug.


By Cyrille Berger at Mon, 2007/05/28 - 5:00am

I wrote that a whole application *should not* be written entirely in Python. I *never* assumed "an application written in python/ruby will be entirely written in python/ruby".

In practice, there is a big risk (and very often an outcome): *a few* slow parts remain in Python although they should be implemented in assembly, C, C++ or whatever fast. I do not know how to state that more clearly.


By Hobbes at Mon, 2007/05/28 - 5:00am

Pages