Re: Getting to Topaz (Was Re: getting on a longer release cycled)

["BJörn Lindqvist" <bjourne gmail com>]

On 9/11/06, Havoc Pennington <hp redhat com> wrote:
> Maxim Udushlivy wrote:
> > I remember somebody compared Gnome with a car. But the desktop is an
> > environment, so it is not a car, it is a parking. The same goes about a
> > hammer: desktop environment is a collection of tools. Different tasks
> > require different collections. The items that you mentioned may fit very
> > well into one desktop ideology (e.g. simplicity) as several profiles.
> >
> > It is possible to make a parallel with Eclipse IDE which has profiles
> > (they call them perspectives). There are profiles for Java source code
> > editing, SVN browsing, debugging, etc. Every profile has its own layout
> > and a set of opened sub-windows (hammers). All profiles are Eclipse-style.
> >
> > Desktops have so-called workspaces (never used them), may be they could
> > be extended into task-oriented profiles?!
> >
[SNIP lots of about GNOME:s focus]
> Still, the broadest, most general-purpose description of IBM Workplace
> is still tightly focused on corporate office workers with IT staff
> (GNOME has not narrowed down to that) and the broadest, most
> general-purpose description of the Eclipse IDE is that it's for
> developers (GNOME has not narrowed down to that either).

I think you are 100% right and that it is important for GNOME to
narrow its focus. For example, if GNOME limited its focus to computers
with 256 MB of RAM, then there wouldn't need to be huge debates of the
advantages and disadvantages of Mono. It would be a no-brainer to
include it because with that much memory, a virtual machine hurts no
one. On the other hand, if it was decided that GNOME targets computers
with less than 64 MB of RAM, then it would be a no-brainer to exclude
it (along with Python and all other memory hungry stuff).

IMHO, it would be better for GNOME to target the high-end machines
because the low-end "market" is already occupied by XFCE.

And it is not just the memory requirements for GNOME that needs to be
decided. I agree that choosing a specific target niche would be very
useful. Problem is, how are you going to do it? GNOME doesn't have a
BFDL (http://en.wikipedia.org/wiki/BFDL) so WHO would decide what the
target niche is? Many of the most successful free software projects
(Linux and Python for example) have a BFDL, that person has a clear
vision about how they want their product to be. Everyone else has just
to accept that vision or leave the project. GNOME is different in that
regard, different developers have different visions and when they
clash, big debates erupt on this mailing list.

Debates that really doesn't solve the problems and doesn't find a
common ground...

It appears to me that GNOME is in many ways in a similar situation as
Debian/Ubuntu (see: http://www.markshuttleworth.com/archives/56 and
http://mjg59.livejournal.com/66647.html). GNOME itself isn't very
focused but has provided others with components on which they can
build a more coherent system
(http://www.novell.com/products/desktop/). I wonder if that clearly
superior Novell-Windows Start menu look-a-like will ever be included
in GNOME?

--
mvh Björn