Re: About writing new apps from scratch

hi Tristan;

On 17 February 2014 04:39, Tristan Van Berkom <tristan upstairslabs com> wrote:

It's a good thing we have some people around who are willing to ask
the hard questions.

not to detract from Sébastian's email (which was absolutely welcome),
but it wasn't really a "hard" question; I've had to answer that
question various times on various social networks and at conferences,
for instance, and I'm not even part of the engagement team.

To answer your concern, this is precisely why we have our module
proposal period as a very important part of each release cycle.

you mean "feature proposal", which is what replaced the module proposal of old.

This means that, while there are always a hand full of alternatives,
basically only one "video player", "photo manager" or "music player"
or whatever is ever officially endorsed by the GNOME community as
the "official" one.

that's not really the goal, and it's also one of the reasons why we
don't do that any more. true, we do have some applications that we
consider "core" — i.e. we can assume they exist whenever GNOME is
installed cleanly, and that fit in with our design tenets — but that
does not preclude, nor hinders, the development of new applications
that can be "endorsed" by GNOME. as a rule, we actually "endorse"
applications that are free and open according to their license, and
that respect and integrate inside the GNOME platform API and design.
we want to celebrate the diversity of the free software world, and
improve our platform for third party developers and integrators, not
subsume branding and communities into our own.

This helps to inspire a spirit of competition - so side projects,
such as the ones you mention (GNOME Music and Photos) - can always
challenge and propose their modules as official again and again.

that's not really how it works, and you and I have *vastly* different
recollections of what the module proposal period did in the 2.x era.

true, we ended up having "official" modules, but I remember conflict
and non-decisions, especially when it came to "controversial"
applications and frameworks. we ended up alienating part of the
community, while still being hated on both sides.

on top of that, seeing applications as "modules" prevented doing
complete integration and usability work; all applications were treated
like silos, and only a certain amount of lip service was given to
accessibility and usability, as well as documentation. a couple of
months are not enough to judge the impact of a new application for our
user base.

Also, through this transparency we attract lots of developers
and contributors - GNOME is a community run project, decision
making is not done by some factions in hidden corners and then
conveniently "announced", but out in the open, every six months
during our beloved module proposal period here on d-d-l.

again, that's a vastly different recollection from mine, which I think
is tinted by nostalgia of a golden age that never really existed.

the feature proposal period is a more vertical process, where
*features* are implemented in terms of their effect on platform, core
user experience, and applications. the feature proposal period is
public, and discussed on the mailing list and on the wiki. it's vetted
by the release team, and it demands that somebody takes up ownership
and responsibility for each feature.

if you have ideas on how to improve the amount of discussion on
desktop-devel-list during that period, I'm all ears.

From here the conversation can go in a few directions - have we
reached the maturity where we can constructively discuss how to
bring back module proposal period and hopefully regain some of
the competitive atmosphere and transparency which was a truly
defining characteristic of GNOME, and I believe a driving factor
in the success of our community in the previous decade ?

to be fair, I kinda resent your implication that there is no
transparency and competition — just because stuff does not happen on
mailing lists any more it does not mean things are less transparent.
Bugzilla is still open; the wiki is still open, the IRC channels are
still open. the Foundation sponsors hackfests, and we have GUADEC and
the Summit.

there are truly no cabals: people can literally show up one day on IRC
(or at a hackfest, or on the wiki) pick up one of the designs, and run
with it — and then present it to the larger community. just like it
happened before.



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]