Seb wrote:
> A database I'm handling is becoming a bit large'ish (~ 30 tables), and
> I'd like to break them down into their natural units. Schemas for each
> of these natural units seems logical, but are they really meant for
> this? I'm also worried about how this would affect programs like
> Libreoffice (the sdbc driver in particular)?
I think that 30 tables is too few to be splitting into schemas based just on
their number; you should have other reasons for splitting them. Also, 30 is
quite small, or at least medium-small; some databases have hundreds, thousands
or tens of thousands of tables.
Think of a schema like a programming namespace. In a program, you may have
different libraries or classes that each contain functions and such. You would
logically group functions together at some times and separate them at other
times. You can let similar concerns organize your schemas, where as schema is
like a library as a function is to a table ... or some other database object
like a stored function.
-- Darren Duncan
