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

[Malcolm Tredinnick <malcolm commsecure com au>]

Aargh!  It's like I'm in a different universe here. My mail supports the
process but presents an alternative view and I get back this bunch of
"corrections". :-(

On Fri, Mar 21, 2003 at 12:24:20PM +1100, Jeff Waugh wrote:
> <quote who="Malcolm Tredinnick">
> > At the end of the day, somebody (or a group of somebodies) have to make a
> > decision. That lot has fallen to the release team, since coordinating the
> > release is really their one true purpose. I agree that the slightly
> > undemocratic nature of that group is of some concern and that when some of
> > the more vocal release team members speak, it is unavoidable that people
> > look at their comments as representative of current thinking (since it's
> > about the only insight we get into that).
> 
> I disagree: We don't make "real" decisions until we're well into the release
> process. We don't decide the modules list. We *do* summarise and finalise
> the discussion, and we *do* make judgement calls once the release process is
> underway.

Which is I said using different words (see the phrase "at the end of the
day"). *shrug*

> It is highly democratic in that whoever puts forward their view will be
> heard. They may not see a lot of agreement, but they will be heard. If the
> release team made decisions at odds with the casual consensus of the
> community, we'd be kicked out of what is meant to be a caretaking/service
> role, swiftly and without remorse. It hasn't happened, because that's not
> what the release team does, or is about.

The composition of the release _is_ undemocratic (which is what I was
talking about, not the process), but that does not mean they are doing a
bad job or the current makeup is unreasonable.  Selection to the release
team, duration of membership, etc. are all done with minimal input from
outside. But, let me repeat, that is _not_ (at the moment) something
that needs to be changed. Please don't read more into my words than is
there, but do read what I said.

[...]
> > All that being said, a lot of these decisions about "what goes into the
> > desktop" are pretty irrelevant. The only benefit of being in the desktop
> > (albeit a large one) is the increased publicity (along with the increased
> > pressure to get it right).
> 
> It's certainly not irrelevant. It's an incredibly important part of our
> decision making process, because vendors and users alike will look to these
> modules as 'best-of-breed' and 'standard'.

For the *particular audience you are targeting* (you cannot talk about
"best of breed" without referring to the target audience, since there is
unlikely to be a universal "best"). For those of us in other audience
groups or working to support other audience groups, what goes into the
desktop release is only as important as everything else, because it is
selected using different criteria to what we are using for our goals.
There should be no debate here, since it is inconceivable that people
think there is only one audience group using GNOME.

> Committing to the desktop or platform release means committing to
> regular releases, project-specific requirements (i18n/a11y/etc), etc.
> It is a commitment to the long term use and viability of these modules
> that users and vendors can trust. It's not a small deal.

This is responding to something I did not write.  Commitments to all of
the features you mentioned can come whether or not a project is part of
the desktop release.

For Heaven's sake: my mail was _supporting_ the current process for
making up the desktop release! However, it was also pointing out -- in
deference to those for whom that release is not appropriate -- that it
is not the end of the world for them if the desktop release is not their
particular cup of tea.

I have spent a very long time over the past few months trying to decide
how my interests fit in with various things GNOME-related. In the end, I
decided that I should just quit worrying about things like the desktop
release, since it is basically unimportant to my use of GNOME as it will
never ideally fit with the way I work. That will not stop me from
working on the developer side of things and helping others over on that
side, since that is where GNOME (an inappropriate collective noun,
really, since it does not differentiate between platform and desktop
targets).

I was trying to share some of the viewpoint with the original poster as
a mitigating factor for why the current process is not a disaster and
why we don't need to keep revisiting old decisions over and over again.

> Yes, technical users who know what they're doing can install whatever
> they want. But when we present "GNOME", we're presenting a coherent
> platform and desktop environment made up of these trusted modules.
> That's important.

In some situations, yes. I completely agree.

Malcolm

-- 
To succeed in politics, it is often necessary to rise above your
principles.