This is kind of off topic in a sense, but what the heck...
Rickard Öberg wrote:
> The test went well. I used Sesame 1.2.3 as repository and used XStream
> to serialize our Java objects to XML and then used the object id's as
> subjects, "urn:data" as predicate, and the stringified XML as object.
> Worked like a charm, with equal or better performance than the previous
> persistence store (which was a serialized binary database, Jisp, works
> kind of like BerkeleyDB).
>
> The next step for me is to index various XPath's in the serialized XML
> as RDF properties of the object, and then all of a sudden we can do some
> very very very cool queries on our entire object model, not to mention
> handle versioning of object models trivially (by using
> "urn:data:<version>" as predicate).
I have now tried to import a database with the website of a reasonably
large customer containing ~300.000 objects. It worked very well. I am
happy with the size of the database (1Gb), and after having indexed some
XML XPath's (hence making the RDF database contain ~3.000.000
statements) we can now do some SerQL queries on the data that we have
not been able to do before in a convenient way. I'm very very happy with
the results so far.
It looks like this is going to be very useful in a real-world
application in a way that I have seen no other data management
technology do before :-)
/Rickard
--
Rickard Öberg
rickard.oberg_at_senselogic.se
_at_work +46-(0)19-173036
Received on Thu Jan 19 2006 - 07:46:06 EST