A little late into the fray (digest subscriber) but I think this is the meat
of the debate. Linus made his opinions clear. I think GGI needs a plan.
Let me offer an example of said plan. Some of this may already be done.
I'm not implying that this is not the case.
1. Make sure you have a very concise design. It should be fairly iron
clad. If you have not thought out a direction completely BEFORE
implementation, bloat tends to set in. Avoid it at all costs. Be simple,
elegant, and targeted. KGI and EvStack are separate ideas. Separate them
completely.
2. Make sure you have detailed design docs that express your goals and
implementation. Your goals should be tightly targeted. "A graphics
interface for Linux" does not qualify. If your goals take up more than a
few lines of text, you are probably not focused enough. You need to say
exactly what you want to do and leave the marketing (I'll convert you to
GGI) stuff out. IMHO, that stuff has done more to set people against GGI
than anything else.
3. Make sure the code you have matches the design docs. I have seen too
many implemention that look like they were written without knowledge of the
design. It makes people look foolish. More fodder for people who already
want to hate GGI.
4. Make sure your patches patch and everything compiles cleanly (again,
more fodder,) and make sure everything is implemented as options. KGI
patches should be one package. EvStack should be a different package. Both
should work independent of the other.
5. When you are ready and confident and 2.3 dev has started, submit both
sets of patches to Linus with the docs. Take time to listen to and answer
his questions. Linus does not hate you guys and if he sees what you are
trying to do, sees the code, and sees that it is The Right Thing, I
personally have no doubt that they will go in as options (as it should in
such a case.) If he thinks you are going in the right direction, but doing
it in the wrong ways, he may not include your patches, but he will probably
tell you why. Don't let egos get in the way. Listen and maybe you can fix
things and they will be included later.
6. If the patched make it into 2.3, listen to criticism. Some people have
some really good points. Some don't know their asses from their elbows.
Take the time to understand who is whom (some people with clues sometimes
appear not to at first glance.) Don't get conceited. If someone offers to
support Trident cards and GGI thinks it is "too good" to support such a
crappy card, then GGI "is not good enough" to be supported by Linux, IMHO.
Egos can kill interest in things quickly. There is not place for it here.
On a personal note, I don't have a single problem with X or have a need for
GGI or EvStack. The ideas are interesting, but ideas don't amount to much
in the real world. I have my X using my Wacom tablet as a pointing device,
I use WindowMaker and Enlightenment, and I have GIMP. I am content. As
long as that stuff will continue to work with GGI, I'm happy to see it
included (iff it buys other people advances AND does not CAUSE problems.)
If the GGI code proves to work my stuff better (over the network as well,)
I'll back it, but I have to see it. Words and ideas don't mean much here.
-Eric ;-)
--
Eric Princen
Micro Analysis & Design
http://www.maad.com/~eprincen
"I don't want to achieve immortality through my work.
I want to achieve it by not dying."
--Woody Allen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu