Re: Gtk/Gnome release schedules

["Dmitry G. Mastrukov" <dmitry taurussoft org>]

В Срд, 05.02.2003, в 07:56, Havoc Pennington написал:
> On Tue, Feb 04, 2003 at 11:43:19PM -0500, Jody Goldberg wrote:
> > On Tue, Feb 04, 2003 at 06:55:20PM -0500, Havoc Pennington wrote:
> > >  a) GNOME 2.4 in 6 months, GTK 2.4 in 9, GNOME 2.6 in 12, GTK 2.6 in 15
> > >                            ^ (with filesel/menus/toolbar/combo)
> > > 
> > >  b) GNOME 2.4 in 9 months, GTK 2.4 in 6, GNOME 2.6 in 15, GTK 2.6 in 12
> > >                            ^ (punting everything but filesel)
> > 
> > > Your mail almost seems to suggest:
> > >  c) GNOME 2.4 in 12 months, GTK 2.4 in 9, GNOME 2.6 in 18, GTK 2.6 in 15
> > > 
> > > if you meant that, that one seems worse than a) or b) to me. In
> > > particular I see no advantage to it vs. a) - it basically just skips a
> > > GNOME release we could make in the summer, for no good reason.
> > > 
> > > So I dislike c).
> > I had not realized that (a) or (b) were even on the table.
> > Both are preferable to (c).  My prefernce would be (a) because it
> > gives us the biggies sooner.
> 
> Yes, my understanding is that the initial/status-quo proposal was a), 
> and Ettore suggested we consider b).
> 
> People's objection to a) is of course no filesel in GTK for GNOME 2.4.
> 
> With these two options the filesel is minimum 9 months out and maximum
> 12, so I'm not sure it should be the deciding factor.
> 
> The only way to get filesel in 6 months will be a), plus some
> cut-and-paste code or private library. I mean, we could try for 
> "GNOME 2.4 in 6 months, GTK 2.4 in 6 months" but I bet that would end
> up just being b).
> 
IMHO GNOME minor version should follow GTK minor. I know, they are only
numbers :) But it's more clear to have GNOME 2.4 based on GTK+ 2.4 and
so on. GNOME 2.4 can be named 2.2.50 (do you remember October GNOME?)
since it's based on the same base toolkit as 2.2.
My second IMHO GNOME should be 2 or 3 month later corresponding GTK+ to
have all its benefits. And a) seems good to me in this aspect,
feature-rich base toolkit is better than semi-feature-rich :) After that
we can have same release cycles with 3 month gap.

Regards,
Dmitry