Re: GNOME: lack of strategic roadmap

[Martyn Russell <martyn lanedo com>]

On 24/02/10 10:11, Murray Cumming wrote:
> On Wed, 2010-02-24 at 09:03 +0000, Martyn Russell wrote:
> > At some point you have to clean up your code base, that's been the
> > case
> > in every project I have worked on. I don't think it is a bad thing
> > that
> > GTK+ is released just "more cleaned up", but others disagree and want
> > 3.0 to have x, y and z major new features.
>
> The problem is that you'll need another ABI break to do major
> refactoring. GSEAL() alone won't be enough, even if it's an initial part
> of it. GSEAL should be part of refactoring, not a reason to release.

I think it is important to do releases when you have progress in the 
project not just because you have some new shiny feature to give to 
people. For 3.0 I can see why you want to have *something* more than a 
cleaner code base of course but I quite like the idea of a GTK+ which 
feels much more solid. I suppose this comes down to if you think 3.0 
should have the sort of changes 1.x->2.x had or not?

> Now you've done the GSEAL() work then we could do bigger work in a
> branch before releasing an ABI breaking release (as stable) that gives
> people nothing but the expectation of another future ABI break, meaning
> that it won't be used much anyway.

Of course. But an ABI break is always better than an API break and if 
recompiling is all that's really needed, the effort by the developer 
linking with GTK+ is really quite minimal (compared to the 1.x->2.x work 
that was required when I ported all my apps back then).

--
Regards,
Martyn