On 09/12/10 17:24, Benson Margulies wrote:
Folks,
We did an evaluation of tuple stores and developed a tentative
fondness for Mulgara. Due to issues with the baroque way in which it
is packaged, I'm having doubts. So I thought it was a good time to ask
you all what you think. The goal here is to store and query a lot of
tuples organized as a lot of graphs.
--benson
(Obviously I'm biased :-)
There is a new Jena-based database/server that is trying to be easier to
use.
Fuseki
http://openjena.org/wiki/Fuseki
It comes packaged with TDB for the database (SDB is also possible).
It is SPARQL 1.1 -- query, update and HTTP update. There are some
command line scripts to drive it and some web pages as well. If you are
loading a lot of data (10's millions of quads), it would be better to
bulk load TDB offline.
It support RDF datasets so it has many-graph support.
Download the zip file: fuseki-0.1.0-VER.zip
from
http://openjena.org/repo-dev/org/openjena/fuseki/
Do note that its very new and although it reuses a lot of mature code
from Joseki, I can't guarantee it's as stable.
It would be helpful to know what you liked about Mulgara. Its good to
learn what makes for ease-of-use.
Andy
