followup to yesterday’s post on GNOME 10.0×10.0

There were some good comments on yesterday’s post; I was going to respond there but I figure they deserve more visibility than they’d get buried in my comments.

I should note that I should not have said ‘GNOME 3.0’ yesterday. GNOME 3.0 implies that we need a successor to GNOME 2.x; I don’t think that is the best way to think about the next GNOME. It might be better to say that GNOME 2.x is Enterprise Desktop GNOME, just like Maemo is (approximately) Handheld GNOME and OLPC is (approximately) Education GNOME. What we probably need is ‘radical user experience improvement GNOME’ in parallel with all those other efforts- GNOME 10.0 x 10.0 is a joke, but you get the idea.

I agree, but more practically the choice of an appropriate BDFL [for the next GNOME] is between what, 6-10 people? –John

The choice of the BDFL is between whoever does the work to generate new ideas and new code. If you have a list of 6-10 people already in mind, you’re looking at whoever has, in the past, generated code or ideas. This isn’t to dump on those people- they have by definition done great work, and good leaders- but leadership for The Next Generation has to come from those doing the work in the Next Generation. Might be the same people, but my guess is that it won’t be, and we certainly can’t know who they are until the work has started- we can’t appoint them beforehand.

Wouldn’t be a start to agree on a GNOME 3.0 mission statement?–Quim

That is exactly the kind of group-planning stuff that Kathy predicts makes for boring, predictable products. The mission statement for what I should call GNOME NG (the relationship with GNOME 2.x (aka the Enterprise Desktop) is tenuous at best, so 3.x is the wrong term to use) is ‘create a radically new and improved user experience’. You can’t plan it much more than that. If it can be planned, it isn’t radically new; it is still an extension of the Enterprise Desktop. I’d be happy to be proven wrong on this one, since it relies on serendipity instead of planning- but I’m pretty confident I’ll be right.

There shouldn’t be this continual hand-wringing about how we get to 3.0. We should be strong enough to say that right now, 3.0 doesn’t matter.–Alex

So, I like GNOME 2.x. I think it is really useful for a lot of people; we likely have at least a million users of various stripes, probably more. But I don’t think we get to 10×10 (or anything like it) without radically changing the user experience. We just don’t have a persuasive reason for Windows users or Windows developers to switch without radical change and improvement. (My posts in early July have some thinking along those lines, and in April of 2005 I listed some potential directions experimenters could go.) So yeah, I think a new gnome software paradigm (almost certainly in parallel with, not as a replacement for Enterprise GNOME, aka 2.x) is important for the long-term health of gnome, the community and the project. (The overarching project for Enterprise GNOME, OLPC, Maemo, etc., being ‘make free software the primary computer interface of the masses’, more or less.)

In short, most people’s current visions can be implemented using the available libraries. –Rob

Almost certainly. Given the continued support Enterprise Desktop GNOME (2.x) will have for many years to come, I’d expect to see something like what the maemo folks are doing- occasionally having temporary forks to work out differences in functionality, but generally working together to enhance a core platform.

There is one area where this might not hold true, or at least, where a large amount of work needs to be done- the competition/integration with (or perhaps cooption of) web apps. See my post from July on this. This problem absolutely must be solved for our platform to have any chance at long-term relevance, and the solutions might be radical enough that they might require substantial divergence from the Enterprise Platform. (Solving the problem might, for example, suggest just wholesale adoption of XUL, or something similarly radical.)