On Mon, 2005-05-23 at 19:08 -0400, Luis Villa wrote: > On 5/23/05, Owen Taylor <otaylor redhat com> wrote: > > On Mon, 2005-05-23 at 12:17 -0400, Luis Villa wrote: > > > Most non-free standards would specifically exclude implementation in > > > licenses we would consider acceptable, no? > > > > > > Luis (who thinks writing more proposals about gnome3, as opposed to > > > going and coding examples of the damn thing, is likely to be a waste > > > of time, sorry) > > > > Let me just put this point of view out in advance of GUADEC: > > > > The only form of "gnome3" that makes sense is to take a gnome-2.1x > > and call it 3.x. > > > > We know from experience that long, incoherent release cycles are > > a mistake. We know from experience that reimplementing code to > > new technology just because that new technology is cool is a mistake. > > > > We have no reason at all to break ABI/API in any fundamental way. > > ABI/API breakage has by itself *no* end user benefits so should > > have nothing to do with the major version of GNOME. > > I don't think anyone except some loons on the wiki is suggesting this > route (though admittedly cleaning out the API to make ourselves more > ISV-friendly doesn't seem too insane.) The serious suggestions, at > least the ones I've seen, are all about radically reworking the user > model. Well, who's a loon and who isn't is very much a question of perspective. :-) Even for major user interface changes, I think that gradualism makes a lot of sense. If we want to rewrite the panel to be filemanager shelf, do that. If we want to make your IM buddies floating semi- transparent heads, do that. Vision of where we want GNOME to go is a long-term thing, not a one-release thing. If GNOME 3.0 is a 2 year radical change of the entire user interface, then what do we do in the 10 years after that? A major user interface shift ever 2 years can't be right. Unlike some people, I don't think the core parts of GNOME are "done", but at some point, work that concentrates on managing existing types of data in Nautilus and the Panel is going to fall into one of two categories: - Really boring maintenance - Rewrite it for no reason or to make a major change in the user interface Neither of these is something that the new GNOME hacker will want to do. In the end, managing files and launching applications is chump-work. The real question is what is in those files? What do the applications do? What can GNOME do for the user? Is GNOME all about the applications now? Should we be writing network servers? Should we be doing mobile device integration? I don't know... these are things we should discuss at GUADEC. To the extent that "GNOME 3" represents the idea that we need to tackle bigger ideas, and maybe tackle things in bigger chunks that make sense to users: great! But I'd fight the idea that GNOME 3 is some major milestone or rewrite. GNOME 3 is just one step on the way to GNOME 3.2 and GNOME 4, and GNOME 4.2 and GNOME 5. Regards, Owen
Attachment:
signature.asc
Description: This is a digitally signed message part