Re: [INTERFACES] [pgaccess-users] RE: bugzilla.pgaccess.org - Mailing list pgsql-hackers

From Bradley Baetz
Subject Re: [INTERFACES] [pgaccess-users] RE: bugzilla.pgaccess.org
Date
Msg-id Pine.LNX.4.44.0207111328310.30839-100000@tomato.home
Whole thread Raw
In response to Re: [INTERFACES] [pgaccess-users] RE: bugzilla.pgaccess.org  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Wed, 10 Jul 2002, Tom Lane wrote:

> Bradley Baetz <bbaetz@student.usyd.edu.au> writes:
> > I'm referring to the mysql |timestamp| type, which will update that
> > column's contents to |now()| when any UPDATE is given for that partcular
> > row, unless the column was assigned to. I don't know how to handle the
> > last part in a trigger.
> 
> It'd probably be close enough to have an UPDATE trigger that does
> 
>     if (new.timestamp = old.timestamp)
>         new.timestamp = now();

Nope, because the documented way of making sure that the field doens't 
change is to use |UPDATE foo SET bar=bar ....|, and thats what bz uses.

Don't worry about this, though - we will hpefully be removing this 
'feature' soon.

Bradley



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [INTERFACES] [pgaccess-users] RE: bugzilla.pgaccess.org
Next
From: "Christopher Kings-Lynne"
Date:
Subject: Re: Should this require CASCADE?