Re: Preparing GConf for the next generation (D-Conf related)

[Alexander Larsson <alexl redhat com>]

On Tue, 2005-03-08 at 13:41 +0100, Philip Van Hoof wrote:
> 
> I've been studying the current GConf code. And now you ask; my
> personal
> view on whats lacking/need to be removed from GConf is the following
> list of items*. Note that it's a personal view. It's my opinion so
> this
> doesn't necessarily means it's the truth.
>
> [snipped long list of implementation details]

Reading a long list of what you think should be done with gconf, or how
you think one could design a configuration system isn't really very
interesting. It just reads like a random list of design decisions, not
at all like a requirements-based design.

What would be interesting would be if someone could spend the time
required to research the actual requirements for a freedesktop
configuration system. Such a system would clearly want to target at
least Gnome, KDE, Mozilla and OpenOffice, so step one would be to study
the systems these projects currently use to learn how they work and what
would be required to replace their current systems. Step two would be to
talk with the developers and users of the current system to try to learn
what missing features they need, and find out the weaknesses of the
current systems (rewriting without learning from past implementation is
useless). Then you can start thinking about the requirements for a new
system design and present for others to comment on. (Note: lowlevel
requirements of the form "should use technology/library foo" aren't very
interesting). 

Of course, during this whole work one also needs to try to build up a
good relation with the people maintaining these systems, because you now
need to try to convince them that your new design is the bees knees.
Just because someone developed something after posting to the xdg list
doesn't mean any desktops will use it. However if you followed the steps
above you stand a good chance of having a system that caters to the
actual needs of these people, and thus you might succeed.

Of course, this is all much more boring than just hacking up something
that you think is cool. I mean, the requirements could end up being
something you dislike or disagree with! The truth is that standards work
is boring tedious work.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Alexander Larsson                                            Red Hat, Inc 
                   alexl redhat com    alla lysator liu se 
He's a one-legged zombie vagrant who hangs with the wrong crowd. She's a 
chain-smoking gold-digging single mother living homeless in New York's sewers. 
They fight crime!