Robert Haas <[email protected]> writes:
> Actually, that's a good idea. But how will you handle tables?
Well, tables are a special case, mainly because it's not clear how to
avoid accidentally throwing away data. (In particular if some column in
the existing table isn't there in the new definition. It's a bit scary
to just drop the column, IMO.) I don't see that that argument applies
to doing an automatic ALTER COLUMN, though, especially since the only
column type alterations that will go through without a USING clause are
reasonably straightforward.
regards, tom lane
--
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers