Re: pg_dump tries to do too much per query

At 12:48 18/09/00 +1000, Philip Warner wrote:
>>
>>You should be able to fix the latter problem by doing an outer join,
>>though it doesn't quite work yet in current sources. pg_get_userbyid()
>>offers a different solution, although it won't return NULL for unknown
>>IDs, which might be an easier failure case to check for.
>
>This sounds sensible; and I think you are right - pg_dump crosses with user
>info relations all the time. I'll look at using pg_get_userbyid, LOJ and/or
>column selects now that they are available.
>

I've just made these changes, and will commit them once I put in some
warnings about NULL usernames - my guess is that this should not stop
pg_dump, just warn the user.

It's a real pity about pg_get_userbyid - the output for non-existant users
is pretty near useless. I presume it's there for convenience of a specific
piece of code.

Would you see any value in creating a pg_get_usernamebyid() or similar that
returns NULL when there is no match?

----------------------------------------------------------------
Philip Warner | __---_____
Albatross Consulting Pty. Ltd. |----/ - \
(A.B.N. 75 008 659 498) | /(@) ______---_
Tel: (+61) 0500 83 82 81 | _________ \
Fax: (+61) 0500 83 82 82 | ___________ |
Http://www.rhyme.com.au | / \|
| --________--
PGP key available upon request, | /
and from pgp5.ai.mit.edu:11371 |/