> > We are annotating nodes on a hierarchical structure
> > where NULL implied an
>
> I don't mean to be rude, but yuck. Why provide a
> record for data that isn't there?
>
No offence taken - I'm trying to improve an old (partially inherited)
system, hence the original post.
I have no idea if th
n
On Wed, Apr 2, 2008 at 8:08 PM, Richard Broersma <[EMAIL PROTECTED]>
wrote:
> On Wed, Apr 2, 2008 at 11:43 AM, Ian Sillitoe
> <[EMAIL PROTECTED]> wrote:
> > I completely take your points - so maybe I should be asking for advice
> on
> > database design instead
I completely take your points - so maybe I should be asking for advice on
database design instead.
We are annotating nodes on a hierarchical structure where NULL implied an
absence of a value in this depth of the hierarchy. As a method of enforcing
this view, we use constraints to make sure we onl
Ah - thanks and apologies for not finding those previous discussions. Does
anyone else feel this might be useful as a point on the NULL section of the
FAQ (it certainly would have saved me an afternoon)?
Cheers,
Ian
On Wed, Apr 2, 2008 at 6:24 PM, Craig Ringer <[EMAIL PROTECTED]>
wrote:
, 2008 at 05:49:37PM +0100, Ian Sillitoe wrote:
> > This is probably a stupid question that has a very quick answer, however
> it
> > would be great if someone could put me out of my misery...
> >
> > I'm trying to JOIN two tables (well a table and a resultset from a
Apologies - just reread my post and I've confused matters with typos during
the abtraction of my code. For the purposes of the example given, please
read "get_colnames_for_id()" rather than "get_cathcode()"...
Cheers,
Ian
On Wed, Apr 2, 2008 at 5:49 PM, Ian Sillitoe
This is probably a stupid question that has a very quick answer, however it
would be great if someone could put me out of my misery...
I'm trying to JOIN two tables (well a table and a resultset from a PL/pgsql
function) where a joining column can be NULL
-- for a given id, return a bunch of col
