lubuntu and the issue of norms by freedesktop.org

[spir]

Hello,

Freedesktop.org [1] is an organism for "open source / open discussion software 
projects working on interoperability and shared technology for X Window System 
desktops"; in other words, they try to define norms on how every bit of desktops 
works (at the interface level), so that software tools & comonents developped 
for one can be used in others, or better that tools can be directly designed to 
work for a whole range of desktops. Freedesktop.org is not an official organism, 
thus its norms do not have the status of standards. Nevertheless, and it's the 
whole point, such norms act or can act as de facto standards that some or even 
all desktop or desktop components may have to follow, if ever they are to be 
used. An independently designed desktop or component, however good it is, may 
end up beeing unusuable for a lack of tools developped for it or a lack for 
interoperability with existing tools; also, users used to some kinds of 
interfaces may not jump to a different one because of its apparent weirdness or 
the cost of learning new practices. Software is inherently monopolistic for such 
reasons (we all know that, they probably are the main reasons why Windows is 
still by far the most used OS).

Recently, i read a thread apparently launched by one developper of lxde asking 
people to take part to discussions at freedesktop.org [2). His point was that 
everything in practice is decided by developpers of the main and big desktops 
Gnome & KDE. This is very bad for lxde in particular because those are about the 
most bloated desktops, which is the opposite of the lightness central goal of lxde.
Thus, ultimately, a question of politics and power, right? Some people suggested 
on the thread to just quit freedesktop.org, possibly to build an alternative 
organism together with user-interface developpers sharing lxde's goals, or which 
goals are compatible. This sounds good, especially if the intention is not 
purely to work against freedesktop.org, but possibly to share a minimal common 
set of norms, or even to ultimately provoke a reaction inside freedesktop.org 
itself which would change its politics and make it possible to follow lightness 
goals while joining back.
But in lubuntu (and I guess many other systems), lxde is used together with the 
Openbox window manager; and Openbox also follows freedesktop.org norms, to the 
point that it is one of its stated goals so that it can be used in conjonction 
with Gnome or KDE [3].
Thus, I guess, a "dissident" politics by lxde, even if conducted together with 
other lighweight projects, would not be able to bring good outcomes; even if 
Openbox leaders would also take part to it. According to their goals, they would 
still have to follow possibly bloated guidelines by freedesktop.org; this, 
unless Openbox happens to never be used in "elephant" desktops but instead only 
in lightweight ones, so that Openbox leaders and decide to center on this usage 
of their project and accordingly change their goals. Another possiblity is a 
fork of Openbox to, say, "Lightbox", but this requires much qualified and 
motivated manpower.

Going back to "share a minmal common set of norms": this may be an alternative, 
inside freedesktop.org, what do you think? There may be a 2-stage set of norms, 
with a lightweight core and a set of optional features. Develpppers of tools and 
component who want them to be usable in lightweight would be encouraged to stick 
with the lightweight set of norms, and avoid optional heavy features.
One problem is where to place the barrier between necessary and optional, light 
and heavy features? The only sensible guideline is, i guess: in case of doubt or 
conflict, leave the feature out; else the lightweight core will ultimately miss 
its goal due to andless & unstoppable inflation (just like programming language 
feature creep); also, once a feature is introduced, it is about impossible to 
removed it even if shows to be a big error a posteriori, because it has become 
used (ditto).
Having this accepted by current leaders of freedesktop.org would certainly 
require a common front of all people whose projects are hurt by these issues, 
and others concerned with them even if it's not their central goal; unless Gnome 
and KDE leaders are more open-minded than they seem to be as presented in the 
pointed thread, and issues were just not well presented enough.

Denis

[1] http://www.freedesktop.org/wiki/
[2) I have lost the pointer, but amjjiwad may give it as he took part to the 
thread, if I remember well.
[3] http://openbox.org/wiki/Main_Page