Re: notice about costly ri checks (2)

On Fri, 5 Mar 2004, Tom Lane wrote:

> Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> > Tom Lane wrote:
> >> I'd suggest something along the lines of
> >>
> >> NOTICE: foreign key constraint "constrname" will require a cross-type conversion
> >> DETAIL: key columns "fkcol" and "pkcol" are of different types integer and double precision
>
> > I suggested the constraint name because of multi-column keys, where he
> > would have to print an arbitrary number of columns in the message. It
> > didn't seem worth doing that work. I see your idea of just printing the
> > column, but that doesn't really point to the primary/foreign key
> > relationship. If the user can't figure out which columns are a mismatch
> > from the constraint name, they have larger problems than this. :-)
>
> Why should we make them guess which column is the problem, when we know
> it perfectly well?

As a side question, if there are multiple cross-type conversions in one
constraint on different column pairs, what do we think the message should
be? One message with multiple column mentions in detail or multiple
notices? (I haven't looked at the patch to see if one or the other is
easier with how it's set up)