You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jena.apache.org by Adrián Quintana <aq...@cnb.csic.es> on 2011/09/28 11:26:28 UTC
How to use different Postgres schemes
Hi All,
We have been testing RDB and SDB. We have understood the general
behavior but we have some doubts. We are developing an Electron
Microscopy Software. We have a base ontology that never change (where
the types and programs are defined) and which is the base of the
application. Each new project in the application will have its own
data/ontology based on it. This project ontology will be update and
modify several times. We want the project information to be stored in
each own schema or table. These new project ontologies will have to be
linked with the base ontology in order to define the different objects
(store the kind of data and the programs that have been used).
We have been looking for scheme support in RDB and SDB but we have not
found any information. Is it possible? How can we do this? If not, has
TDB or any this feature?
Best regards,
Adrián
PS: We have already read the mailing list entry non-SQL databases and
combining SDB and TDB models which explains similar problem but we have
not understand the procedure...
Re: How to use different Postgres schemes
Posted by Andy Seaborne <an...@apache.org>.
On 28/09/11 10:26, Adrián Quintana wrote:
> Hi All,
> We have been testing RDB and SDB. We have understood the general
> behavior but we have some doubts. We are developing an Electron
> Microscopy Software. We have a base ontology that never change (where
> the types and programs are defined) and which is the base of the
> application. Each new project in the application will have its own
> data/ontology based on it. This project ontology will be update and
> modify several times. We want the project information to be stored in
> each own schema or table. These new project ontologies will have to be
> linked with the base ontology in order to define the different objects
> (store the kind of data and the programs that have been used).
> We have been looking for scheme support in RDB and SDB but we have not
> found any information. Is it possible? How can we do this? If not, has
> TDB or any this feature?
> Best regards,
> Adrián
> PS: We have already read the mailing list entry non-SQL databases and
> combining SDB and TDB models which explains similar problem but we have
> not understand the procedure...
RDB is deprecated - it's only there for existing installations.
SDB has a fixed schema for RDF - there isn't a per-application table
design. That's one of the advantages of RDF - you can mix in extra
information later without needing to reorganise the database.
If you want to have one project - one RDF data store (for isolation?),
then you could either have multiple SQL databases.
TDB similarly is a generic RDF datastore. Not being SQL, it does not
even have "tables" in the SQL sense. It has multiple indexes.
Mixing and matching RDF data from different stores is done by having a
general purpose DataSource object and adding the models from different
stores.
Andy