Re: Outstanding patches

Philip Warner <pjw@rhyme.com.au> writes:
> ie. The CHECK constraints inherit only at the time of table creation. I
> think this is a bug in  ALTER TABLE for CHECK constraints.

More like an "unimplemented feature" ;-).

After thinking for a moment, I believe the only real gotcha that could
arise here is to make sure that the constraint is adjusted for the
possibly-different column numbers in each child table.  There is code
available to make this happen, or it might happen for free if you can
postpone parse analysis of the raw constraint tree until you are looking
at each child table.  Just something to keep in mind and test while
you're doing it ...
        regards, tom lane