Re: Removing pg_migrator limitations

Bruce Momjian wrote:
> I wasn't aware enum ordering is something we tried to maintain.
> One issue is that we are not supporting the addition of enum values even
> for people who don't care about the ordering of enums (which I bet might
> be the majority.)
>

The ordering of enums is defined and to be relied on and I think it's
absolutely unacceptable not to be able to rely on the ordering.

We should never be in a position where the values returned by
enum_first(), enum_range() etc. are not completely deterministic.

Part of the original motivation for implementing enums was precisely so
that they would sort in the defined order rather than in lexicographical
order. It's a fundamental part of the type and not an optional feature.
The idea of potentially breaking it makes no more sense than allowing
for a non-deterministic ordering of integers.

cheers

andrew