Re: plpgsql: numeric assignment to an integer variable errors out

Bruce Momjian <bruce@momjian.us> writes:
> Whatever happened to this patch?

I think we bounced it on the grounds that it would represent a
fundamental change in plpgsql behavior and break a whole lot of
applications.  People have been relying on plpgsql's coerce-via-IO
assignment behavior for ten years.  If you prefer coerce via
cast conversion, you can get that by writing an explicit cast.

Now it is true that a lot of the uses for that were subsumed when
we added coerce-via-IO to the native cast capabilities; but I'm
still quite scared of what this would break, and I don't see any
field demand for a change.
        regards, tom lane