Re: Issuing a press release about GNOME 3

["Peteris Krisjanis" <pecisk gmail com>]

2008/7/18 Thilo Pfennig <tp pfennigsolutions de>:
> Luis Villa wrote:
> > In short, I think you're letting minor technical considerations (and
> > perhaps perceived pressure from KDE?) set out an agenda, rather than
> > making the user and improvements for the user set the agenda, and I
> > think that is exactly backwards, screwing us up with users,
> I was thinking about the same but hesitated to write that as I am used
> to having a non-majority view. I think one could also think about the
> KDE disaster as that they did a simlar thing: 4.0 was not really the new
> big thing for the users - but rather a quite stable new API. The
> difference in GNOME  3.0 would be that although the motivation to do a
> higher number comes from the ABI/API changes but it is expected to be
> MUCH more stable than KDE 4.0.

KDE 4.0 should have been called "KDE 4 developer release". Having
stable API and core features is nice and stuff, but it really don't
sell anything, if nothing works.

> But still thats one of the core problems that marketing follows function
> and not the other way around or at least on the same level. So marketing
> still is seen as a bunch of people who spread rhe word of new
> functionalities and versions.
>
> What GNOME should have is a good general strategy where it is heading.
> Right now there are only some random "feature clusters" like the Online
> Desktop or the mobility stuff. But thats not really an outlined idea or
> vision.

Problem is that it is really hard to lay out any concrete version with
such diverse community as GNOME. RedHat cares for Enterprise, Ubuntu
cares for noncommercials/oems/home users, Novell cares for
compatibility with Windows.

What do you actually seek is concrete feature set, aka specification.
However, again, within open source community, it is very hard to make
concrete spec and fullfill it - and for obvious reasons - there are
ever going lack of dev power, lot of stuff is boring to be coded (for
example, everyone is ready to play around with Clutter, but Gnome Scan
still lacks developer power. Needless to say wich I would like to
working in first place), supporting and maintaining stuff is serious
business.

However, AFAIK, it is possible to define current spec - features,
things we have for now - and track progress (or regress) from here. It
is also can be used as compare base for next (aka 3.0) GNOME version.
Do we have callendaring program for 3.0? Check. Do it has CalDAV
support as now? Uncheck. Bug. Task. Processing :)

It is needed, no matter how hard it is. Because it is next level -
when we actually start to care what happens after code. Being
proactive about features and problems about them is way forward.

> Dact is that GNOME was never build from a users perspective but from a
> developers perspective and also from the perspective of distributions.
> If it should take the users into account the users have to have a role
> in the development process - like having a users council which is
> involved when new releases are planned. I think to expect developers and
> distribution to takes the view of the user is maybe futile, because they
> will have their own view and interests.

That is a long stretch - in fact, distributions listen to user input a
lot (Ubuntu has UbuntuForums, Brainstorm page, cool bug system), and
their work is pushed upstream. There maybe is huge problem how to
coordinate upstream stuff, because lot of features are delayed for
some unknown reasons.

> One could try to select a good mix of volunteers from different
> backrounds  who are willing to test new features and give feedback - or
> who say a word about ideas that are floating around. and they could be
> heard by the Foundation board and maybe have one representative in that
> board. It would then be good if they have a setup which allows them to
> test new stuff without being technical experts.

Testing should be done, yes, but again, nor GNOME nor distributions
don't have such manpower. Afaik installing betas of next greatest
distro version is enough. What is not enough is bug reporting (Only
Ubuntu has made it with ease), commenting, submitting info. For
example, Evolution doesn't work with some kind of strange POP3
servers. How get such information and debugging info from user? How to
submit changes and then encourage user to test them? That is question
should be asked.

Of course, mainstream oriented distros should look for more casual
testing (user groups and stuff), but it is up to them.

As far as I see, GNOME has good foundations to be dominant free
desktop environment. And it can be easily achieved building upon what
we have already, doing incremental changes, making them more directed
and planned, following spec.

Just my two euro cents,
Peter.