[
Lists Home |
Date Index |
Thread Index
]
> There's nothing built in. Dave Winer et al are proposing OPML for this
> purpose because it's hierarchical. So instead of a ton of pointers in
> the RDDL, you'd have one, to the OPML doc.
>
> > 3. Going back to Tim's initial proposal, are there
> > any thoughts on what that
> > "small (but extensible) RDF vocabulary" would look like?
>
> I'm beginning to think RDDL is just the ticket for this. -Tim
A deeply hierarchical system strikes me as being potentially problematic for
this, in that there would be dependencies running from root to leaf. For a
big site I think this could be a pain to maintain, especially if the site
wasn't organized in a hierarchy itself. On the other hand I can certainly
imagine a bunch of RDF files acting as a very shallow hierarchy, each file
containing a bunch of statements that would (apart from the references
between these files) be completely independent of each other. Hopefully a
half-decent RDF store would be able to work on the liddle-biddy RDF
statements as a cohesive whole when required.
Assuming we don't want a single site-wide metadata file, how would you
picture the RDDL data being distributed - shallow tree again? Or perhaps a
single root RDDL file would distribute its 'payload' amongst a bunch of RDF
(or even OPML) files?
Cheers,
Danny.
|
