Re: Bidirectional Binary Self-Joins

On Apr 1, 11:35 am, "David Cressey" <cresse..._at_verizon.net> wrote:
> "Marshall" <marshall.spi..._at_gmail.com> wrote in message
>
> news:1175415285.221875.155230_at_n76g2000hsh.googlegroups.com...
>
>
>
> > On Mar 31, 10:26 pm, "David Cressey" <cresse..._at_verizon.net> wrote:
> > > "Marshall" <marshall.spi..._at_gmail.com> wrote in message
>
> > > > > Unfortunately we are left with that same issue - the choice of team1
> > > > > and team2 is arbitrary, and as David or Paul pointed out, it means I
> > > > > have to do some jiggery pokery on my queries to consider both these
> > > > > different attributes simultaneously, even though they are playing
> the
> > > > > same roles. This obviously isn't going to kill us, but it doesn't
> seem
> > > > > elegant, so it niggles.
>
> > > > Get over it. :-)
>
> > > What if, instead of soccer, the game is Chinese checkers, where there
> can
> > > be up to six contenders in one game? Or how about tournament bridge,
> with
> > > say 30 pairs competing in a match?
>
> > > You could end up writing an awful lot of "or" expressions.
>
> > Yeah. At six you're likely looking for a new answer. But by that time
> > you're probably not identifying games by the capricious collection
> > of competing contingents.
>
> I found myself looking for a better answer with regard to so called "vital
> statistics", namely births and marriages, where there's more than one
> principal participant.

The situation seems to occur generally in any relation that has a candidate key composed of two or more attributes playing equal roles (or principal participation as you put it). And those attributes always has to be a subset of the antecedent of any material implication in the propositions being collected.

However I am in the process of getting over it ;) Received on Sun Apr 01 2007 - 16:43:16 CEST