Re: Modulesets Reorganization

[Lucas Rocha <lucasr gnome org>]

Hi Paolo,

2010/6/2 Paolo Borelli <pborelli katamail com>:
> On Wed, 2010-06-02 at 00:37 +0100, Lucas Rocha wrote:
>> The long term plan for the GNOME applications that were removed from the
>> Desktop, Admin and Dev Tools modulesets is to simply highlight the high-quality
>> applications using the GNOME platform through our communication channels
>> (release notes, website, etc). There will be no "official" apps anymore and no
>> 'Applications' moduleset in the GNOME releases. The goal here is be more open
>> with the app developer community around GNOME and to highlight all the nice
>> things that can be created using our platform.
>
>
> Personally I think the idea of giving up on having an Applications
> moduleset is a very unfortunate decision for multiple reasons:
>
>  - first and foremost it will drive away fresh forces from gnome: most
> of new developers get involved in writing code for applications and
> feeling "officially part of gnome" is a determining factor in getting
> more involved in desktop-wide hacking, in the community mailing lists,
> join the foundation and so on
>
>  - Cross-application hacking will slowly vanish, slowly ruining the
> quality and consistency of the gnome desktop. For instance until a few
> years back we had ui revisions for all gnome application each cycle
> which ensured consistency and quality... I think we should work on
> getting this kind of thing back (especially since gnome 3 will bring
> some changes to the ui!) instead this decision will make each
> application go its own way
>
>  - QA effort will be impacted: I used to run a full jhbuild version of
> gnome all the time and I don't anymore and I think I am not the only
> one... This is a problem on its own (and something the RT should work
> on!), but if "jhbuild build" does not build a full desktop anymore (or
> if it takes 2 days to build because it pulls in 6 media players) we are
> once again moving in the wrong direction
>
>  - A particular note goes to also giving up on the Dev Tools moduleset:
> new developers ask all the time about which tools should they use to
> hack on gnome, the dev tools moduleset was starting to become a decent
> answer to that (though I certainly wish it provided a more complete
> development environment and more tools were added, eg sysprof). Giving
> up on creating an official moduleset is a step back since we cannot say
> "Want to hack on gnome? Here are the tools of the trade"
>
>
>
> I am really sorry to say it, but this feels like the gnome chronic
> inability to make decisions: we are not able to pick RhythmBox or
> Banshee, we are not able to deal with Zeitgeist being on LaunchPad,
> etc... so the only decision we are able to make is to cop out on making
> decisions.

I don't see this as a move to avoid decisions. It's more about being
inclusive with app developers. Why do we have to choose between
high-quality competing apps? If they follow GNOME guidelines, use our
platform, are well implemented, have a lively user community, I'd
prefer to have both considered as first-class citizens in GNOME.

I would only care about not having competing implementations of the
same thing in desktop core which is the part where we actually want to
have more control.

--lucasr