Re: Date, Darwen, Pascal and the alternative to Nulls in the RM

From: JOG <jog_at_cs.nott.ac.uk>
Date: 23 Mar 2006 18:40:50 -0800
Message-ID: <1143168050.603283.183180_at_v46g2000cwv.googlegroups.com>


falcon wrote:
> JOG wrote:
> > Paul Mansour wrote:
> > > Assume one accepts, as I do, the argument against nulls put forward by
> > > Date et al. Would it be fair to say that at this point in time they
> > > really don't have a solution to missing information?
> > > anything better. Am I wrong here?
> > [huge snip of interesting stuff]
> >
> > They propose a system that decomposes all relations down to irreducible
> > tuples (hence eradicating nulls). The horrendous joins and
> > consequential problems this might leave are handled by their closed
> > source "transrelational model (tm)" in a manner which currently eludes
> > my recall. As far as I know this system is yet to see the light of day.
>
> So their solution to the null problem is to convert n-ary relations to
> binary relations? If that is the case then that is what I was thinking
> in my previous message.
>
> As far as I know, transrelational has to do with how data is stored on
> disk, not how it is logically represented...no? If we ignore the
> problem of 'horrendous joins,' don't we still have the issue of nulls
> showing up when we 're-construct' a tuple with joins (as mentioned
> previously by another poster)?
[snip]

Yes my point exactly. And when this null appears again you have a proposition in your resulting relation (if you are still calling it that) that does not satisfy the given predicate. Theoretically this hardly appears to adhere to the mathematical and logical rigour that RM proponents claim for the system. Pragmatically its nothing one can't handle. Nonetheless it is food for thought.

> [111-11-1111, jim, 21, A]

Sadly, I am no longer 21. In fact I can safely say a null would have been preferable.
All best, Jim ;) Received on Fri Mar 24 2006 - 03:40:50 CET

Original text of this message