On Tue, Feb 20, 2007 at 04:02:00PM +0100, Leonard Ritter wrote:
what exactly is the problem?
i think there are two separate cases: revolutionary next-gen
systems, and shorter term evolutionary improvements.
In the first case, the main problem seems to be that no one
has any good ideas. At least not good enough to break
I'm happy to see a time of relative stability. People can
really get to know the system, push it to its limits and
actually write some decent applications.
I would have thought that any big change would have to
involve OpenGl graphics hardware. (And by this I mean
enhancement of existing 2d drawing functions, not
walkthrough desktops). Progress is being made here at the X
and window manager level, and once that is stable, maybe
it'll be clear what the next step is. On the other hand,
proprietary graphics acceleration could all turn out to be a
In terms of evolutionary improvements, I'm happy to see
things happening slowly but surely, and the mistakes of the
past being corrected.
One major shift is the redoing of functions that were
implemented incorrectly at the "Desktop" level. Positive
developments include cross-toolkit IPC with dbus, the
replacement of gnome-vfs with one at the glib level, the
disembowelling of proprietary desktop soundservers, etc.
In terms of Gnome, well, should we just come to terms with
the fact that Gnome is a disaster area? Perhaps its demise
will have the benefits of bringing us such goodies as a
general purpose "Display Manager", general purpose session
manager, mime icon system, application configuration system,
more variety in the WM area etc. Unless these functions are
devolved - as difficult as that may be in the short term -
developers will continue to be frustrated at the difficulty
in penetrating the inevitable wall of conservatism and
indecisiveness necessary for the stability of such an
established monolithic pile of cards. Perhaps Gnome was
always doomed? Lets face it, it doesn't have the gloss to tie
it together as a finished product - in fact it doesn't even
have the basic apps. It doesn't have a native browser, (due
to the problems of Mozilla architecture) it doesn't have a
file manager (Nautilus is not a file manager), and it doesn't
have a nice terminal. And inevitably it doesn't have the
"finishers" - the people with the most important yet
time-consuming and unrewarding job for a desktop intended
for non-geek use - the artists and technical project
He he, maybe the above paragraph is a little harsh. I think
there's a point in there somewhere though :-)
In terms of Gtk, it does seem that one of Gtk's problems is
that it is "good enough". Its a very powerful system, even
if very frustrating in some areas. No doubt Gtk-3 will have
an improved layout system, but its a really difficult task
to come up with a system that can satisfy everyone without
it being too big. At the moment the people with the both
the vision, and cross-discipline technical expertise to
carry off a major and worthwhile api change don't seem to be
in evidence anywhere.
The loss of some funding from Red Hat will presumably mean
that minor upgrades get delayed. I'm thinking of things like
deciding on which of the already-written gtk canvases (such
as Goocanvas) to use as a replacement for the old
GnomeCanvas, additions to the theming system, better html
support etc. (Perhaps even a themable rotary widget for us
audio guys!) Meanwhile, composite widgets continue to get
added. While the fiasco over the file dialogue, for example,
is frustrating, the basis is there for someone to step up
and improve it.
And maybe I should just mention Freedesktop.org. Some good
work being done there (some funded by Red Hat, even). And
lets not forget the many other much improved libraries such
as cairo and popplar.
Is Eugenia a troll? Sure. I think she's right in this case,
but that's no bad thing. I got a bit disturbed after reading
her "wake-up call", but after some reflection, I feel
positive. Know your strengths!
Audio is a another matter, but that's a different discussion