Daily Archives: May 3, 2006

New KDE pages online

As Janet Theobroma mentioned two new sites around KDE have been launched: KDE-artists.org and open-collab.org.
Both are relaunches of old ideas/websites which already existed for quite a time, but needed to be refreshed since the aims, possibilities, but also because the knowledge of how to address and present things and thoughts has been improved a lot over the time. So what is it all about?

KDE-artists is, like the name says, the place for everyone in the KDE project how comes into contact with art of any sort: if you are an artist, the new site will be probably the site where you spend most of your time, speaking with other artists about what and how to do stuff, what tools you use, which nice and useful tips you have, and so on. Of course the same is true for everyone who wants to become an artist inside of KDE: kde-artists.org will provide you with information, tutorials, standards, themes, guidelines, but also with contacts to other artists or for example tools with which you can work to develop art.

The other new site, open-collab.org, the sucessor of the old Kollaboration forums, is, as the name indicates, completly about collaboration: if you want to share some thoughts about any kind of application – do it there! Especially if you are an end user and want to discuss some stuff, for example together with some mockups this is the right place. You do not have to be a programmer or such to write there, the opposite is the case: open-collab.org is a plattform for people who want to share ideas because they want to see thm but are not able (due to whatever reasons) to implement them. It is also a place to share some good ideas about which you think everyone should know.
On the other hand, if you are a developer, make sure that you keep an eye on that page since it is perfect to contact users and see what they are thinking – feedback is good and can help you keeping an eye on what the users want or dream of. And you can use this forum to contact motivated developers who do not have a project right now: write what you would like to see implemented in your project and what is easy enough to entry the project as a newbie, enthusiasts will read it!
There are also some analysis of Gaps and problems in the FLOSS world documented, and this is something you shoul keep in your mind also – for example because you want to improve your app in this field or you want to prepared for discussions about FLOSSoftware.

And, btw.: the collaboration is not KDE only, one of the main mentioned programs there is inkscape :)

OpenGL and Canvas in KDE 4

There have been some posts in the last days about some upcomming graphical features in future KDE applications.

The first is about OpenGL functionality in KOffice: after Krita already started with implementing some OpenGL features into the program to speed up working with graphics Ariya Hidayat now started to hack some code into KOffice’s KPresenter to give it some nice effects. A blog post gives a nice impression what we can expect in the next versions of KOffice.
I really appreciate this because I think that KOffice is the long term solution for KDE, and probably for the FLOSS community at all since I do not like the performance of OpenOffice at all on a Linux machine :-/

Another graphics improvement we will see in KDE 4, or better, in Qt 4.2 using applications, is GraphicsView (or sometimes referred as QGraphicsView which is the name of one of the main classes of the project GraphicsView). GraphicsView is a replacement for QCanvas which is the standard canvas class of Qt. For the non-developers: a canvas is a 2D area of a program where you have items like texts, polygones, circles, animated pictures, etc. It provides the base field for all interactive/dynamical elements of a program.
So this part will be improved now, and this improved version will be the base for KDE 4.
And although I do not know which features QCanvas has, the list of features for GraphicsView sounds promising:

  • All items can receive key, mouse, focus, and accessibility events.
  • Items can now be rotated, scaled, etc., individually.
  • Support for OpenGL.
  • Animation support, not by static item velocity.
  • The text item has as powerful editing capabilities as QTextEdit.

You will find more in a QDevBlog post, together with some nice screenshots. I have to admit that these screenshots calm down my fears concerning the new graphic subsystem of Vista a little bit: I once saw some demos of the new subsystem, and this demo had similar features: animated, 3D accelerated buttons and similar stuff. That was very impressing and I lost a bit hope that Linux can battle Vista in an appropriate way. But when KDE 4 takes use of these new features, KDE/Linux does not have to fear anything :)

If you want to see a little bit more of GraphicsView in action, look at this blog post from Zack Rusin. He also describes some abilities and features together with the usual screenshot ;)

WLAN in Linux – current state

Today announced Devicescape that their Linux WLAN solution is Open Source and will be integrated into the official Linux Kernel. This is true, but you should know that this release was not today but some months ago (nevertheless the step is still a nice step :) ). To clearify the situation of WLAN in Linux the technical news magazine heise.de has published an article surveying the whole topic – if you can read at least some german, go through it, it is worth it!

For those who are not able to read german, here is a short summary:
The reason why they announced it today is that Red Hat’s John W. Linville, the official manager of the WLAN subsystem, announced in an e-mail that this stack will be the future of WLAN networking in Linux and that most of the WLAN driver developers agree on this topic. bcm43xx based cards and some of the Ralink-WLAN-Chips chips are already supported.
There are no statements about the old IEE80211 subsystem and the SoftMAC extension, so we can estimate that they will be dropped quite soon.

Other problems are that the ipw3945 drivers are still troublesome since Intel had to take account of local laws and therefore the driver needs an userland application which sets the features according to the local law.

Nevertheless we will see a common standard WLAN stack for Linux in the future which will speed up development and improve the driver quality a lot. But there is still quite a lot of work to do, especially in the cases where the drivers depend on proprietary stuff.