Sam Carleton wrote:
The server of my client/server program is currently using SQLite, an
embedded file based SQL engine. I am looking for a client/server
based RDBM and have narrowed it down to either PosgreSQL or Firebird.
I would prefer to use PosgreSQL, but the management of the physical
files are concerning me.
A very key element of the program I am developing is the ability for
the user to pickup one folder that contains all the data for that
"document" and copy it somewhere else, then allow the program to read
it from that other location. It is equally important that when the
database server initially start, if one of the "tablespaces" was
deleted by the user, the database be none the wiser.
This is, of course, not a problem with Firebird since it is file
based. What I could do is simply disconnect from the tablespace each
and every time the server program shutdown and reconnect to that
tablespace each time the server starts up, pointing to a new
tablespace if need be. So how would one:
1: "disconnect" from a tablespace, in such a way that the tablespace,
nor it's contents is changed.
2: "connect" to an existing tablespace.
I don't think PostgreSQL is going to work for you if thats a
requirement. A tablespace doesn't contain the metadata for the items
in it, thats stored in the main cluster space (pg_catalog, etc)
--
Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
