[postgis-users] 1.3.3 Pre-Release Notice
Paragon
lr at pcorp.us
Thu Apr 10 18:46:18 PDT 2008
Mark has been hanging around databases too long which always do cost based
short-circuiting of some sort. They don't follow the same rules of conduct
as other languages and don't even follow each other.
E.g. I think PostgreSQL will always resolve = and other operators etc first
before functions regardless of order and then go left to right (8.3 adds a
twist since you can put in a hint as to the cost of a function to change its
behavior). This has bitten me on several occasions when I forgot that
sometimes left to right is the rule of the day and sometimes it isn't.
SQL Server does some random ordering which seems to follow no rhyme or
reason at times although it does seem to slightly favor right to left
short-circuiting. On very bizarre occasions it refuses to short-circuit at
all when it obviously can.
And of course great old VB.NET gives you an option - because maybe sometimes
you don't want to short-circuit.
E.g. you really want to carry out a costly call do
a And b or c
If you want the expensive b and c sometimes but short-circuit when you can
do
a AndAlso b orElse c
Thanks,
Regina
-----Original Message-----
From: postgis-users-bounces at postgis.refractions.net
[mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of Paul
Ramsey
Sent: Thursday, April 10, 2008 6:23 PM
To: PostGIS Users Discussion
Subject: Re: [postgis-users] 1.3.3 Pre-Release Notice
My intuition that it worked left-to-right came from Perl... where'd your
reverse intuition come from, Mark? (Languages are so
multifarious...)
On Thu, Apr 10, 2008 at 3:26 PM, Mark Cave-Ayland
<mark.cave-ayland at siriusit.co.uk> wrote:
> Fascinating. Being self-taught in C, it's always interesting to learn
> something new - even if it does go against intuition :)
_______________________________________________
postgis-users mailing list
postgis-users at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users
More information about the postgis-users
mailing list