"Christopher Kings-Lynne" <chriskl@familyhealth.com.au> writes:
> No - that was for my pg_stat_reset patch, and at the time of submission it
> didn't conflict with any other oids...
Yeah, things are moving so quickly now that patches which add catalog
items may easily hit each other in transit. The best short-term
solution seems to be to expect committers to run 'duplicate_oids' before
committing anything that adds catalog entries.
In the long run we might want to think about reassigning the
hand-assigned OIDs in some more-rational fashion; the current OID layout
is a complete mess with no visible rhyme or reason...
regards, tom lane