I would expect if I did intcol =3D 4.35 or intcol =3D 'abc', it would throw=
a
type mismatch error.
Now whether you want to be clever and make intcol =3D 4.0 not throw an error
and instead convert it to intcol =3D 4 is another thing. In that case, I
still wouldn't scan the table. And now that I think of it, 4 is different
than 4.0 in terms of precision. If you are querying an intcol, maybe that
doesn't matter, but I would probably err on the side of precaution and throw
a type mismatch error as well.=20=20
Gabriel
_________________
Gabriel Weinberg
yegg@alum.mit.edu=20
-----Original Message-----
From: Stephan Szabo [mailto:sszabo@megazone.bigpanda.com]=20
Sent: Wednesday, April 07, 2004 12:58 PM
To: Gabriel Weinberg
Cc: 'Bruno Wolff III'; pgsql-bugs@postgresql.org
Subject: RE: [BUGS] Sequential Scan Index Bug
On Wed, 7 Apr 2004, Gabriel Weinberg wrote:
> Presumably, but that is not what I was doing.
I was responding to the part that was: "I would expect it to throw an error
immediately, instead of scanning the table for a value of a different type."
If say intcol =3D 4.345 is an error, is intcol =3D 4.0 an error as well giv=
en
that 4.345 and 4.0 are presumably the same type? I'm not sure what error
you would expect.