See the attached text file. It didn't want to paste well. Some of them are more ideas than set-in-stone goals, so keep that in mind.
Thanks, Donnie
Gentoo desktop goals for 2005
=============================
=============================
Leads: Donnie Berkholz, Brandon Hale
Overall goals
=============
- Keep the "stable" desktop current
- The vast majority of users reportedly run stable
- Pmount
Current Gnome desktop systems have facilities for auto-mounting
removable media devices, utilizing udev, hal, and dbus. A potential
improvement to this is using pmount [1], a wrapper for mount that allows
normal users to mount pluggable devices in a more secure and sane
manner. Two bugs [2], [3] are currently open on Gentoo Bugzilla with
more info on what is needed to move forward with this. Once this meets
with foser's stringent QA, I think it would be a welcome improvement
to the Gentoo Gnome desktop.
[1] http://lists.debian.org/debian-devel/2004/11/msg00201.html
[2] http://bugs.gentoo.org/show_bug.cgi?id=69575
[3] http://bugs.gentoo.org/show_bug.cgi?id=69577
- Network Management
Include some kind of simple, GUI-able network management application.
This is especially important to mobile laptop users, who often find
themselves moving between a number of wired and wireless networks with
unique configurations. How are other distros addressing this common
problem?
Red Hat is developing NetworkManager [4], a wide reaching protocol stack
covering several aspects of network configuration, as well as a
notification applet to control the active network connection. The latest
release seems buggy and not completely functional, and CVS seems to be
regressing rather than moving forward from a stability and
maintainability standpoint.
SuSE has developed netapplet [5], another notification applet aimed at
letting mobile users easily associate with various access points or
connect to wired networks with the click of a mouse. This one currently
works fairly well on an Ubuntu system. Gentoo might possibly need
some more work to get going; it appears [6] our own Obz has already
taken a crack at this.
[4] http://people.redhat.com/dcbw/NetworkManager/
[5] http://primates.ximian.com/~rml/netapplet/
[6] http://www.livejournal.com/community/gentoo/128030.html
- Streamlined Boot Process?
There is current interest among other distributors in shortening the
time from power on to login in their default install. The efforts were
kicked off with the development of bootchart [7] in response to a
challenge [8] by Owen Taylor. Some progress has been made with
readahead, X, and gdm by Daniel Stone, and mentioned in his blog [9].
Whether or not anyone from Gentoo takes interest in joining this work,
we should continue to "watch this spot" and pick up on improvements in
this area.
[7] http://bootchart.sourceforge.net/
[8]
http://www.redhat.com/archives/fedora-devel-list/2004-November/msg00447.html
[9] http://www.fooishbar.org/blog/
Subproject goals
================
Video
-----
Lead: Luca Barbato
Overall:
- mantain the status quo
Some more or less important things to be addressed:
- Global cleanup to some cflags (me).
- Update some tedious ebuilds and make them work with the new X11 layout
(ChrisWhite, me and the others)
Games
-----
Lead: Mike Frysinger
- Add more games, fix more bugs, and have fun
- Chris Gianelloni is working on GameCDs in collaboration with the release
engineering TLP
Accessibility
-------------
Lead: Deedra Waters
- Maintain the accessibility packages
- Trying to port speakup to other arches to be included on LiveCDs. It works on
amd64 and x86; we're working on sparc and ppc.
Research
--------
Lead: ???
- Needs a new lead to jumpstart it
- Good opportunity for someone who wants to be involved in turning Gentoo into
a prime desktop distribution
GNOME
-----
Lead: Marinus Schraal
You'll have to ask foser. He's playing his cards close.
KDE
---
Lead: Caleb Tennis
I don't know if we (KDE) have any cohesive goals, per se, but a few things
come to mind:
1. Introduce split ebuilds into portage. This will allow people to emerge
only the KDE packages that they want. These will be going into portage
very soon, riding on the release of 3.4_beta1.
2. Integrate Qt4. I'm currently working on this, and I expect others will
jump on the bandwagon as soon as it becomes a little more popular. Right
now, it's very buggy, and nightly builds don't even build for me. It's in
portage, and we continue working on it. This ebuild should be FHS
compliant, as well.
3. Get upstream library dependencies fixed. When emerging KDE, it may
sometimes pick up libraries it finds without any explicit "--with-??? or
--without-???". This is bad, because portage doesn't bind KDE to those
libraries, so they can be uninstalled without knowing KDE needs them. We
plan to start tracking these and getting them fixed upstream.
Science
-------
Lead: George Shapovalov
Well, even though we are heavily strained for resources (developer time, but
who in Gentoo isn't :)) Scientific Gentoo fares along quite well. We just
completed a category split, so the immediate goal would be to monitor how it
goes for a little while more and, as we accumulate even more packages, to
split off a few more categories. Oh, to recruit more people as well of
course. Some researchers are becoming so adventurous that they actually
propose help maintaining their areas of interest :). Now that I am becoming
more active again I'll issue a few more calls for help, that should get few
more people in..
Other goals:
finalize fortran.eclass;
complete transition to new versions of blas/lapack ebuilds;
Adjust involved packages to use new eclass and blas/lapack libs;
regular maintaince.
Desktop-WM
----------
Lead: Brandon HAle
Only 12 bugs open here, nothing very high priority, but it would be nice
to work towards 0 in the new year. A few of these don't even look
relevant to the member herds. Several of the WM's here are dead or very
slow moving targets upstream, and are relatively bug free, so since a
small cleanup blitz at the onset of this project, the overall bug count
has been fairly low. A bit of room for improvement, but nothing pressing
here.
Controversial issue:
Having been on an extended absence from Gentoo, SuperLag has agreed to
solve any problems that crop up with Openbox, and ciaranm still
faithfully maintains Fluxbox. Blackbox, however, lacks an interested
maintainer for the aging WM and its numerous accessory packages. These
aren't in much distress, but they aren't getting any newer. Ciaran has
mentioned several times that we should consider removing this. Possible
user outcry, we need a way to gauge use somehow.
X
-
Lead: Donnie Berkholz
Core X11 will continue moving away from the /usr/X11R6 FHS exception and toward
standard FHS compliance. This is supported by several X maintainers from other
distributions, and a bug is open in the FHS database to drop /usr/X11R6. The
move should be done in Q1, perhaps Q2 at the latest. This will leave a single
symlink /usr/X11R6 -> /usr.
Drivers are doing quite well, especially the more popular ones such as ATI's and
Nvidia's. We have close contact with ATI's dev team, and we'd like to get closer
with Nvidia's over the next year. Bryan Stine has been doing a great job with
getting drivers such as Wacom and Synaptics building against the SDK and has
become our de facto SDK expert, so we may see things happening with the SDK.
Andrew Bevitt and Luca Barbato have been primarily maintaining the Nvidia and
ATI drivers, respectively, and I expect to see their excellent work continue.
Andrew's also been improving opengl-update quite a bit, and we expect to see our
OpenGL switching become even more seamless and user-friendly over the next year.
We could really use more people to help with the core, because I've got less and
less free time. I've had a recruitment effort going for at least six months
without a whole lot of luck. Adam Jackson, an upstream X dev, has been helping
out a lot with bugs. Also, Adam Mondl is helping with hardened X.
There are a ton of open bugs, and I'd like to get those lowered by quite a bit.
Desktop-Util
------------
Lead: Brandon Hale
The goal of this herd when I founded it was to clean up the huge mess
in x11-misc and get things up to date as possible and then put in a bit
of time every so often to keep things from deteriorating. I believe we
largely succeeded at this. However, our herd became a dumping ground for
bugs that don't seem to fit anywhere else, particularly new ebuild
requests. There are currently 86 bugs assigned to desktop-misc@, and a
vast majority of them are ebuild requests for packages of limited
usefulness, or functionality that is provided by several other packages.
These need to be reviewed, and hopefully closed. If someone is
interested in maintaining a package here, great, but we don't need to
*add* unmaintained packages to x11-misc. It would be ideal if the
wranglers would not dump bugs here.
Goals: Possibly get one or two fresh faces (ideally already developers)
to volunteer some time to poke around the bugs and x11-misc and clean
house. Don't expect them to stick around very long, I've experienced
first hand the high potential for burnout due to the huge mass of ugly,
uninteresting packages here (not that I am the most dedicated guy
around).
signature.asc
Description: This is a digitally signed message part
