Roy,

> Goody - where's the definition published please.
SQL 2003 Standard / Foundations / Pages 60,61
(page numbers are for the draft documents, since I don't have the final
documents)


--
Fernando Dias
[Team Elevate]

Tim,

> No, it isn't.  You're attempting to posit that something is known when it
> isn't known. It doesn't matter what the mathematical rules are for a certain
> operator - the result is unknown if one of the operands is unknown.


Except in the special case of "<Unknown numeric> x 0" ...which is known
to evaluate to zero irrespective of the other operand.

Eryk

Tim


You're not supposed to be here. You said you weren't playing anymore

Roy Lambert

Fernando


Is there a url for that or do I have to pay money?

Roy Lambert

Eryk,

<< Except in the special case of "<Unknown numeric> x 0" ...which is known
to evaluate to zero irrespective of the other operand. >>

I understand that as the theoretical result, but as far as I've read in the
standard (6.26 Numeric value expressions), it isn't part of the standard.

--
Tim Young
Elevate Software
www.elevatesoft.com

Tim,

> << Except in the special case of "<Unknown numeric> x 0" ...which is known
> to evaluate to zero irrespective of the other operand. >>
>
> I understand that as the theoretical result, but as far as I've read in the
> standard (6.26 Numeric value expressions), it isn't part of the standard.

I agree, your reading of the standard is correct. I was merely citing it
as an example where the SQL standard sacrifices logic for intuitive
consistency. Logically "NULL x 0" cannot evaluate to NULL for exactly
the same reason that NULL AND FALSE must always evaluate to FALSE.

Eryk

Eryk


I obviously need a new intuite <vbg>

Roy Lambert

Roy,

You can find it here:
http://www.wiscorp.com/sql_2003_standard.zip

--
Fernando Dias
[Team Elevate]

Fernando

>You can find it here:
>http://www.wiscorp.com/sql_2003_standard.zip

Brilliant. I'm downloading it for those sleepless nights.

Roy Lambert