Joerg Hessdoerfer <Joerg.Hessdoerfer@sea-gmbh.com> writes:
> PG has *some* views in the system catalog, which make life easier, but
> some essential(?) things like 'list all tables in DB' has to be done
> in a multi-table join with special attributes. What is the rationale
> of that? Wouldn't it be easier (and more portable, see 7.3/7.2 system
> catalogs vs. psql) to have views for that?
Only to the extent that the views match what a particular front-end
actually wants to see.
Peter Eisentraut is currently working on adding the SQL-spec-mandated
"INFORMATION_SCHEMA" views; so as long as all you want to know is what's
in the spec, those should be your answer. But I do not foresee psql or
pg_dump ever switching over to INFORMATION_SCHEMA, because they want to
know about some things that are Postgres-specific.
regards, tom lane
