Ibex Release

I was talking to my father (an old school programmer) last night and I 
mentioned all the work I tried to do on XWT/Ibex over the lecture 
break.

     "Ha, I've been hearing about that thing for years
      and you still haven't got anything to show for it."

He's right; I've been working with all of you on Ibex on and off for 
years, and we still don't have anything. Apparently, it isn't even 
ready for designing widgets, let along applications.

When will Ibex be usable?

Today? Tomorrow? After one more feature? It seems to me now we've 
skipped XWT 2.0, and are going straight to XWT 3.0. Are we ever going 
to say, "no more commits to core.ibex.org unless their bug fixes"? Are 
we ever going to say, "the reference doesn't change without a core@ 
vote"?

As I see it, core.ibex.org is ready for feature freeze. True, it 
doesn't have SVG. The concept of a template still isn't completely 
separate from user interface design. But it's bloody good. Good enough 
to be better than all the alternatives, and good enough for me.

As I see it:

- Immediately branch core.ibex.org. Only bug fixes go into it from
   here, all other feature development happens in a different repo.
- Freeze the reference document as is with the core branch. Any
   changes need to be voted on by [email protected]
- Bug fixes must be tested *by the person writing the fix*,
   against a stable widget demo.
- Put all effort into a stable core release with the current
   feature set.
- Remove the web site we have now, put up a few static pages.
   Give someone access to update it.

I consider this much more important than everything we've been doing 
for the last few months. As I have said before, Ibex has become more of 
a cathedral than Hurd. At this rate, it will be ready just in time to 
be its Window Manager.

The only difference between Ibex and Hurd is no-one's ever heard of 
Ibex.

After my mid-semester on Friday, I'm going to start working on 
Wildebeest. If we haven't started coherent work on a stable branch by 
the time I have it working, I'm going to fork the project. And instead 
of spending hours every day trying to keep it in sync with 
core.ibex.org, I'm going to spend the time fixing bugs and making it 
usable for application development.

I really want to see this idea succeed. While the incredible techniques 
use to achieve the goal are fascinating, they are meaningless to me if 
it doesn't do anything.

-- d