[fdo-users] PostGIS : new feature class naming proposition

Mateusz Loskot mateusz at loskot.net
Tue Feb 12 13:48:38 EST 2008

Bruno Scott wrote:
> We are actively debugging the postgis provider.
> And i have a proposal for the featureclass name
> Actually , on the connection string we specify the datastore.
> As this datastore is phycally the schema name so there is no need to include
> it in the featureclass name


yes, this is correct.

> So i suggest to get rid of this schema name but to keep backward
> compatibility

I assume you mean compatibility between current solution and further 
improvements adding multi-schema support, right?

> So if someone use an fdo function having a featureclass name as an input
> parameters both "schema~table" or "table" will work
> but if we ask the provider to give us the list of featureclass name
> it will only return "table"
> Is that ok?

Yes, I think this is a very good idea.
Generally, the featureclass and featureschema names are used to uniquely 
identify a physical table in datastore.
Considering we are going to support multiple PostgreSQL schemas and
multiple FDO schemas/classes, this solution should be sufficient.

Thanks Bruno!

Mateusz Loskot

More information about the fdo-users mailing list