Thoughts from GUADEC
Frederico’s talk had some bits that reminded me of a “Topaz” desktop design that came up on desktop-devel-list back in ‘06. It wasn’t given much time of day back then though. Check it out.
The concept of the computer as Personal Assistant appeared in a few forms this year. Concepts like it helping you organize your information and time and managing your attention. Its obvious there’s been a few threads of thought on this bubbling away for a while now and this year you could see ideas starting to crystallize. Very cool!
Its obvious we’re going to need some good query capabilities over large RDF trees. To my simple mind this seems to basically involv unification over tuples, so prolog keeps popping into my head. Has anyone checked out SeRQL?
Tags: computer as PA, crack, GUADEC, Topaz
July 12th, 2008 at 11:13 am
Hi,
I’ve been hacking on beagle as a gsoc project and then moving to the l3s to write my diploma thesis about desktop metadata networks. I’ve mostly been using beagle++ which links the beagle indexing capacities to a rdf store you can query using sparql and serql (the two seem very similar to me). My impression of serql was that it’s nice to have but it lags aggregators. Since i needed them because i was mostly doing statistical stuff i ended up transfering the data into a sqlite db.
Anyway - I don’t really get what you would need serql for. That’s probably because i have not been to guadec. Really looking forward to hear some more about the plans.
July 12th, 2008 at 3:40 pm
Can’t wait till everyone comes back from GUADEC to discuss these things and slides from the presentations come online.
Until then, could you please elaborate on what the advantage of SeRQL is over SPARQL? Thanks.
July 12th, 2008 at 8:56 pm
Oh, SeRQL implents SPARQL as well as its native SeRQL. I was more just pondering if prolog would be a better basis for a RDF store/query system than an relational database. I don’t really know what the pros and cons are. I would like to investigate the literature more when I get the chance.
July 13th, 2008 at 2:59 am
Oh, you were referring to the actual RDF implementation on that page. Looks interesting. I know nothing about Prolog, but I could imagine it would have an easier time with SPARQL (both being declarative languages) than, say, C.
May I ask if any concrete ideas or plans for RDF came up at GUADEC? Did Federico mention it in his talk?