[Mageia-dev] GNOME plans

[Wolfgang Bornath]

2012/6/5 Colin Guthrie <mageia at colin.guthr.ie>:

> You've maybe not been following the Gnome OS stuff then? I have and it's
> very interesting. As you know I've been a KDE developer for many years,
> but I much prefer the Gnome approach to lower level things. People
> involved in the Gnome project are very supportive to drive forward the
> core underlying plumbing layers rather than just build abstraction
> layers on top to rationalise the underlying insanity. I certainly have
> much more affinity to this approach to engineering (as you can likely
> tell when I share my opinions on stuff) than the KDE approach which does
> quite often paper over the cracks rather than solve the underlying problems.
>
> So I don't disagree with you with regards to how we, as a distro, should
> treat Gnome, but do keep in mind that "Gnome OS" as a whole is something
> we will all benefit from - this covers everything from how early boot
> should deal with user interaction and encryption passwords etc (i.e.
> plymouth) right up to GDM and X11 initialisation (including how to
> handle things when it fails) and thus the desktop environment "Gnome"
> will come with certain requirements for the underlying system on which
> it runs. In most (if not all) cases, these requirements will be for the
> general good anyway.

I certainly agree on your first paragraph (cause it makes sense and
the "Gnome way" is promising and it is trying to achieve a Good
Thing(tm)). Except for one point i already pointed out: a good part
(if not the larger part) of the users are not using Gnome. So,
whatever Gnome includes in its settings - if it is something which
affects the whole system (like "everything from how early boot should
deal with user interaction and encryption passwords etc (i.e.
plymouth) right up to GDM and X11 initialisation (including how to
handle things when it fails)") then distributions still need the tools
for these same tasks bnecause of users who do not use Gnome.

If this tool for "non Gnomers" is such a central toolbox like the MCC
a duplication of tools for the same task is unavoidable with users who
use Gnome (as we have now), because you can't stop this system being
installed on the same system as Gnome. Also keep in mind that many of
those tools are needed before the desktop environment is even
installed.

So, what I'd rather see is a common cross-distribution toolbox outside
of the desktop environments taking care of system business than a
common toolbox inside a certain desktop environment which is not
available if you do not use that desktop environment. This has not
much to do with development but more with logic and reality.

-- 
wobo