Excerpts from rikard.pavelic's message of s=C3=A1b feb 25 10:23:18 -0300 20=
12:
> But I would expect second alter to pass and enforcing not null and default
> when adding this column in table and not enforcing not null and default w=
hen
> adding into composite type for another table.
>=20
> Is this by design, oversight or a TODO?
I think this is more a TODO than anything else. Last year we discussed
something similar to this -- twice, even; IIRC, one was buried somewhere
in the discussion about "variant" types, if you want to search the
pgsql-hackers archives. As far as I recall, discussion died mainly
because no one had the time and/or energy to pursue it, not because it
was impossible.
--=20
=C3=81lvaro Herrera <alvherre@commandprompt.com>
The PostgreSQL Company - Command Prompt, Inc.
PostgreSQL Replication, Consulting, Custom Development, 24x7 support
