Re: pg_dump versus views and opclasses

"Brendan Jurd" <direvus@gmail.com> writes:
> On Fri, Jan 16, 2009 at 9:01 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> "Brendan Jurd" <direvus@gmail.com> writes:
>>> * It seems there's no pg_depend entry for
>>> types/functions/operators/opclasses that the view depends on, unless
>>> they are part of the SELECT list.
>> 
>> What PG version exactly?

> This is all on 8.3.3.

Okay.

Looking at it some more, I notice that the SortGroupClause dependencies
are on the individual operators, which probably isn't good enough: the
operator *classes* have to exist or the parser will complain when trying
to make sense of the view.  So that would be a good thing to change for
8.4 (and it's not too late yet).  However --- it's also the case that
pg_dump should dump all operators *and* operator classes before it gets
to views.  So either you were doing something funny with the dump/reload
or else there's a circular dependency in your DB that pg_dump is
breaking in a bad place.  I look forward to the test case ;-)
        regards, tom lane