These are mutually exclusive --- I see no reason to do both.
Not sure that's true; we've taken te design decision to make allow user-defined constraint names to be non-unique. Given that, I think we should allow people who fall into the trap to be able to use the info schemas to get details of their constraints. So, adding enough detail about the constraint to uniquely identify it, even if it is a user-created one, seems essential.
=> Adding table identification info to constraint details in the info schema is necessary.
I don't agree that using OIDs to in constraint names is bad; the table name will be misleading when tables are renamed, and encourage use of internal data (PG_* tables) when info schemas should do the job for most people. I think we're confusing a presentation issue with an internal design issue.
----------------------------------------------------------------
Philip Warner | __---_____
Albatross Consulting Pty. Ltd. |----/ - \
(A.B.N. 75 008 659 498) | /(@) ______---_
Tel: (+61) 0500 83 82 81 | _________ \
Fax: (+61) 03 5330 3172 | ___________ |
Http://www.rhyme.com.au | / \|
| --________--
PGP key available upon request, | /
and from pgp5.ai.mit.edu:11371 |/
---------------------------(end of broadcast)--------------------------- TIP 6: Have you searched our list archives?
http://archives.postgresql.org
