Developer conferences give us a window on the workings of free open source software. The aKademy conference opened with talks on a wide range of subjects, giving us an idea as to the direction of our favorite project. I intend to give summaries of the talks, and I'll start with Daniel Stone's talk about the Freedesktop project. The talks from the first weekend of the aKademy conference are available at http://ktown.kde.org/akademy/. Over the coming weeks I will give highlights of the talks. Jonathan Riddell has also produced transcripts. They are at wiki.kde.org.

The Freedesktop project has 60 projects including X.org, Cairo 2d vector graphics framework, D-Bus system wide IPC, Gstreamer multimedia framework, UIM/SCIM universal input support and HAL platform indpendant hardware abstraction layer.

X.org is the XFree86 fork which is moving to improve X. Daniel talked about some of the goals of the project, such as a modular structure permitting separate releases of modules, and a reworking of XLib.

Cairo is a vector graphics library which supports rendering to X, images and OpenGL. It is already supported as a backend to QT4 and GTK.

D-Bus is a universal inter process communication (IPC) system with system and per-session busses. It will allow deep integration to the underlying system. One example he used is a camera being plugged into USB bus. It will allow the desktop to be notified without the desktop being aware of the underlying system. Using HAL, network configuration could be broadcast, allowing KDE to respond without knowing about the underlying system. It somewhat duplicates what is already in KDE and Gnome.

UIM and SCIM refer to universal and shared common input methods. The differing demands from various languages had created a situation where many wheels are invented to address specific issues. This is an attempt to provide a universal solution.

Freedesktop will soon release its platform for developers.

The question period was most interesting, since there is some distrust among KDE developers of this project. One questioner asked how the decisions were arrived at, suggesting that it looked alot like Redhat. Daniel answered that he and others decided to put together a release. It is up to others to use or ignore the project.

Freedesktop.org is an aggregation project. If a standard is produced it is because the desktops worked on developing it and implementing it.

I don't think KDE has anything to fear from freedesktop.org. On the contrary, there is much to be gained. Already common standards are in place for menu entries and the like. FOSS development always starts with code. Next is where others find a use for the code, then contribute to finish, improve or adapt the code. Same with fd.o. Aaron Seigo mentioned an example where work was being done on the taskbar standard, and he suggested improvements or changes, which were accepted. So if someone doesn't like the direction of something on fd.o, pipe up, send patches, get involved. Or ignore it. My impression from the talk was that most projects come from individuals or groups trying to solve a problem, not some underhanded way to impose an unwanted standard on anyone.

Sashmit Bhaduri announced that Beta 6 of akregator 1.0 has been released.