Well, after reaching a complete impasse of how to proceed with Ibex, we (all the developers bar the main Ibex developer) had to fork.
There's not even been a beta Ibex release (version Nitrogen), only a release under it's old name XWT and that was a year ago (version Lithium). And we were told by the main developer to drop Lithium because Nitrogen was only 2 months away and we should be stress testing it by writing widgets.
Well, 100s of bugs, lots of testing, and years of man hours later the guy says to us, "Nitrogen should not be considered stable enough for writing widgets". WTF? That was the last straw... as if the unscheduled updates, the refusal to answer basic API-change queries, and the refusal to sanction a stable release ASAP in favour of yet more pissing about with the core code, and much more than I care to mention here all just wasn't enough.
So last Monday we forked. A beta release by the end of next week. A stable release in the next 3 months. Lots of documentation, lots of activity, lots of public noise.
It's current (and probably final) name is Vexi.
And just to describe the kind of infuriating tasks we suffered under Ibex, I've spent the last week (a whole week of man hours) tracking down and rewriting bits of one of the main Java classes in order to eliminate some terrible bugs. And this code I'm replacing is code rewritten in the last 2 weeks by said main developer. Not tested, not thoroughly thought through, just imposed on his community. And he accused us of trying to blackmail him when we said, "if you're not going to be more diligent, we're going to fork." Well we have and this is the last time I'll have to mop up his mess.
The guy is a great coder, but he doesn't appreciate people enough and he doesn't understand that for a community to build up you actually have to release software. Releasing the perfect software may be a utopia, but you're better off releasing imperfect but good software than nothing at all.