Re: [Rhythmbox-devel] How about a release ?

[Jonathan Matthew <jonathan d14n org>]

On Mon, Feb 16, 2009 at 11:00:36AM +0000, Bastien Nocera wrote:
> On Mon, 2009-02-16 at 20:52 +1000, Jonathan Matthew wrote:
> > 
> > I hope to be able to make a new release in the next month or so.  This
> > will probably be numbered 0.12.0.  After that, I'm not planning to do
> > much new development work on the rhythmbox project as it currently
> > exists.  I'll still fix (some) bugs and review patches, but it's too
> > much of a dead end for me to do more than that.
> 
> Why is it too much of a dead-end? The feature list stacks up nicely
> against other music players, and although using C might be making our
> life harder in some cases, I'm not sure I see ways to replace it in many
> distributions.

I'm not disputing that at all.  If I just saw rhythmbox as markedly
inferior to a similar project, I'd probably go and hack on that, or let
them continue as they are and find something else to do with my time.

I don't have a concise summary of the problems I have with the current
rhythmbox.  It's been building up over the last six months or so, maybe
longer.

Anyway, my main problem is that the underlying data model is too rigid
and limited:
- the set of queryable entry properties is fixed
- entry properties can't have multiple values
- albums don't exist, they're synthesized from whatever
  entries happen to match the album name
- file metadata, observed user data (play count, etc.), and user provided data
  (ratings) are all stored together when they'd be more useful separate
- the database is designed specifically to support implementation of the
  genre->artist->album browser and doesn't work so well for other
  things.
- it doesn't always make sense that sources are completely distinct from
  each other

These things are pretty much baked into every aspect of rhythmbox.
Rhythmbox is fine for what it is, but I don't think I can make it into
what I want it to be.