Re: SQL spec/implementation question: UPDATE

On Oct 22, 2007, at 11:01 AM, Alan Hodgson wrote:

> On Sunday 21 October 2007, Kevin Hunter  wrote:
>> Heh.  And as Tom points out downthread, that "shortcut" probably  
>> doesn't
>> gain anything in the long run.
>
> Considering how expensive updates are in PostgreSQL, I suspect that  
> isn't
> true.

No, the idea is that the vast majority of updates aren't going to  
have this happen so you're waisting cpu cycles trying to determine  
beforehand if the update will have any effect.  The gain from not  
running updates that actually wouldn't change any table data would be  
dwarfed by the overhead of making that check on every update.

>
> However, the current behaviour does seem to be logical; we did in  
> fact ask
> for the row to be updated ...

Right.  And, as someone else pointed out, it also allows any triggers  
on the table to still run.

Erik Jones

Software Developer | Emma®
erik@myemma.com
800.595.4401 or 615.292.5888
615.292.0777 (fax)

Emma helps organizations everywhere communicate & market in style.
Visit us online at http://www.myemma.com