Monday, November 12, 2007

Feature plans for Fedora 9

Fedora 8 is out now and plans for Fedora 9 are made. The feature list is already in place and contains several entries.Fedora 8 has just been released, but of course plans for the next version are already under way: a Fedora 9 Feature List has been created in the Wiki, just like the one for Fedora 8, and people and groups now add their plans and aims.
Two outstanding aims are the inclusion of KDE 4 as the default KDE version and the usage of PackageKit as an alternative package management frontend. Since the Fedora 9 release schedule aims at May 2008 both aims should be realistic and possible: KDE 4 is supposed to be released in a working state around christmas, and PackageKit already works pretty well now.
Another feature which is currently lacking in Fedora 8 is real support of RandR 1.2: while everything works in the command line there is now GUI yet. Even worth, the current GUI shipped with Fedora is not aware of the new RandR features and asks you to restart your X - which is totally nuts. There are also no fallbacks implemented to always have at least one working output. At least the first part will hopefully be addressed by the next Fedora release.Speaking about broken tools and implementations, NetworkManager is again part of the feature list. The team aims at getting NM into shape that it can be activated by default and just works in all situations: with mobile support via blue tooth, support for more than one device, support for system wide configuration via PolicyKit, etc. Since Fedora 8 ships with a development version of NM 0.7 some of these features might find their way into Fedora 8 as an update as well.
For PulseAudio the developers plan to replace current volume control principles by a PulseAudio volume control. The background is that current systems rely on the information exposed by ALSA which are neither user friendly nor useful for everyone. Therefore a PulseAudio logic will be used to also use the possibilities of PulseAudio.
There are also plans to create a new Fedora Spin for that release: Fedora Astronomy. That plan also contains some Fedora related political “problems” is also “political” interesting: Fedora is quite GNOME concentrated, but most of the astronomy software for Linux - KStars, Xephem, Celestia, Stellarium, Partiview - has little to do with GNOME. But KStars, which is quite well known and is used for example to control telescopes of any size requires KDE libraries anyway, so it would be useful to install KDE as the default desktop. But from my point of view I would not be surprised if something else would happen.
Speaking about GNOME, GDM will see a major rework for Fedora 9 as well. It will be ported to all the new possibilities which are around these days on modern desktops: improved fast user switching support, PolicyKit support D-Bus API, etc. While these features make great progress and are supported by more and more distributions and applications, the underlying desktops still fail to support some of these. I don’t know of any work at implementing PolicyKit’s features deep into the KDE libraries at the moment, for example. And in general D-BUS is still not used in a way it could be in both desktops: you could start entire different sets of applications depending on the time, the (network-) location or the mood of the user (photo database together with the image manipulating application, or sets of audio applications for music work, etc.), and you could also give one save&close command to such sets of apps at the same time with D-BUS.In this regard the rework of GDM is “just another” step, but a necessary and important one.
It shows that Fedora 9 will again come up with multiple changes - many of them are not even listed here. Most of them will be evolutionary (which is not bad!), except maybe for KDE, which will see a revolutionary step to a new level.

No comments: