Re: wierd AND condition evaluation for plpgsql

On Wed, 2002-05-29 at 02:36, Joel Burton wrote:
> > -----Original Message-----
> joel(at)joel=# select true and seeme();
> NOTICE: seeme
> ?column?
> ----------
> t
> (1 row)
>
>
> It certainly appears to be short circuiting for "select false and seeme()",
> for instance.
>
> It appears that this isn't short-circuiting by order of expressions, however
> (as Perl and other languages do); for example, "select seeme() or true"
> doesn't ever get to seeme(). I assume PG can simply see that the statement
> "true" will evaluate to true (clever, that PG!), and therefore it doesn't
> have to evaluate seeme() ?

Are these intricacies of SQL standardised anywhere ?

I know that gcc and other ccs can achieve different results depending on
optimisation level - usually this is considered a bug.

But as PG runs always (?) at the maximum optimisation, should there be
such guarantees ?

Or is it something that should be ind doc's/faq's (- don't rely on side
effects) ?

------------------------
Hannu