Michael Nolan <htfoot(at)gmail(dot)com> writes:
> So, a <> operator (either xid,xid or xid,integer) would need to be
> implemented using the hash opclass, correct?
No, it's unrelated to the opclass. It'd be worth marking it as the
negator of the equality operator, but otherwise it'd really be
unconnected to anything else.
regards, tom lane