We currently don't represent the sequence privileges in the information
schema. We could perhaps do a subset of them.
In the SQL standard, there is only the USAGE privilege, and its only
purpose (AFAICT) is to allow the NEXT VALUE FOR expression.
PostgreSQL's nextval(), by contrast, requires both USAGE and UPDATE
So with regard to the information schema, we could
a) show USAGE privileges as is (and perhaps SELECT and UPDATE as is, as
b) show USAGE privileges only if UPDATE is also granted
c) show nothing, avoiding the whole issue.
pgsql-hackers by date
|Next:||From: Tom Lane||Date: 2011-04-26 20:28:51|
|Subject: Re: SR standby hangs |
|Previous:||From: Merlin Moncure||Date: 2011-04-26 20:19:35|
|Subject: Re: Bad COMPACT_ALLOC_CHUNK size in tsearch/spell.c?|