on crawling upnp contentdirectories...

[Jussi Kukkonen <jku linux intel com>]

Hi all,

I've been experimenting with crawling UPnP contentdirectories
(mediaservers) lately. For those who haven't seen earlier discussion,
the rationale for doing this is that many meadiaservers do not support
search, and the browse trees they offer can be pretty weird -- or at
least very different from the UI design of a specific media player UI.
Crawling and caching would allow us to implement very fast searching
locally, avoiding that problem.

I've got a grilo-upnp implementation of this (see crawler branch):
  http://gitorious.org/jku-grilo/grilo-plugins
  git://gitorious.org/jku-grilo/grilo-plugins.git

...but after some discussions we decided I shouldn't send the patches or
continue the work, at least for now. Reasons for this:
 * caching idea is orthogonal to what grilo currently is
 * there are performance issues -- like memory use -- that could be
problematic on mobile/embedded scenarios. Grilo just isn't designed for
holding unknown numbers of media items in memory "just in case"
 * we should be able to do better than per-application, per-instance,
non-persistent caching...

So, I'm currently looking at implementing this feature in a better place
in the infrastructure. Tracker seems like an obvious choice, even though
that probably means quite a bit more work than working with your nice
little upnp plugin.

There's one commit that you might want anyway: e165b81b removes the
GUPNPAV_OLD_VERSION checks and makes the code a lot more readable. I can
send that as patch as well if that's easier.

Cheers,
  Jussi