Stephan Szabo <sszabo(at)megazone23(dot)bigpanda(dot)com> writes:
>> (We can tweak the RI triggers to make the runtime failure message be
>> more helpful than "Relation 0 not found".) ISTM the point of having
> Are we mostly concerned about the case where it's 0 or all cases
> where the constrrelid relation doesn't open?
I'd only special-case OID 0; we can reasonably assume that came from
a broken CREATE CONSTRAINT TRIGGER definition. Any other failure is
harder to guess at the cause of, so we should just let the generic
error reporting do its thing.
regards, tom lane