On Wed, 2006-10-25 at 14:06 +0100, Jamie McCracken wrote:
> > I do wonder why nobody has taken a fast database-backed RDF triple store
> > (from librdf), put a sparql-based frontend on it for queries (again from
> > librdf) and used re-used a lot of code from both Beagle and Tracker for
> > harvesting the metadata...
>
> Its only an option if we can customise it to fit in with what we have.
>
> tracker is both an indexer with full text search and a metadata DB and
> right now librdf looks only relevant for the metadata DB side of things.
>
> In particular we can in tracker cross search both full text indexes and
> database indexes so we can quickly search for "get me all files that
> were modified in the last 7 days which have a keyword of Tracker and
> contains the word GNOME anywhere in it". Would this still be possible if
> I used librdf?
librdf would give you a W3C "semantic web" compatible metadata store.
Full-text indexing is separate but if you use the same URIs in the both
the metadata store and the full-text index, then I can't see a reason
why not.
Ross
--
Ross Burton mail: ross burtonini com
jabber: ross burtonini com
www: http://www.burtonini.com./
PGP Fingerprint: 1A21 F5B0 D8D0 CFE3 81D4 E25A 2D09 E447 D0B4 33DF
Attachment:
signature.asc
Description: This is a digitally signed message part