Re: 2.4 Module List and Rationale (aka GEP10 and 11)

[Andrew Sobala <aes gnome org>]

On Thu, 2003-03-13 at 18:56, Luis Villa wrote:
> I've put up GEP 10, Standards for Inclusion in the GNOME , as well as
> GEP 11, Module List for 2.4.
> 
> http://developer.gnome.org/gep/gep-10.html
> http://developer.gnome.org/gep/gep-11.html

My comments:

- 2.3.3 GNOME-ness

I don't think we should specify that apps must use "GNOME 2
technologies", but rather should completely interoperate with other
GNOME 2 applications and the desktop or some such wording. Using GNOME 2
technologies is ambiguous and implies that an app _must_ link against
libgnomeui (or other libraries). Some of these such libraries (such as
libgnomeui) should be in Gtk anyway.

I think if a Gtk app operates well with the rest of GNOME with no holes,
there's no reason why it shouldn't be included. Or even if a KDE/Qt app
integrates well with GNOME (although this is not the case with the
current implementation of GNOME and KDE technologies), why it should
necessarily be excluded.

2.3.4 Free-ness

What do we mean by "Free or Open?" Presumably this means licenses that
GNU say are compatible with the GPL
(http://www.gnu.org/licenses/license-list.html#GPLCompatibleLicenses) or
are approved by the OSI (http://opensource.org/licenses/).

2.4.1 Quality

If we exclude repeatable crashers, some core GNOME modules don't qualify
:-). Maybe this should read something along the lines of "Very few (I
mean <=2) repeatable crashes, and the maintainers are dedicated to
fixing bugs like this quickly".

2.4.5 Use of GNOME resources

I think the important bits here are:

-> CVS and translations. How well do the translation team cope with apps
out of GNOME CVS (like gstreamer, IIRC). I think this is the only reason
an app actually needs to be in GNOME CVS.

-> If not in GNOME CVS, should have a public CVS repository. The GNOME
cycle revolves around having bleeding-edge testability.

-> My biased point of view is all modules _must_ be in bugzilla.
Otherwise, forwarding bugs to maintainers is a total PITA.

Extra bit

What about saying that modules must be prepared to conform to the GNOME
freezes in at least one CVS branch.

My thoughts...

-- 
Andrew Sobala <aes gnome org>

"If we eventually have the ubercool component system - based on Bonobo, or
something else - then great, we can then proxy it over IIOP, D-BUS, SOAP,
and morse code." -- hp