Re: [Info-ingres] Re: is there an equivavlent to auto_increment in ingres ?


"Roy Hann" <specially@xxxxxxxxxxxxxxxxxxxxx> wrote in message
news:IqmdnRYslYvwz0PfRVnyhQ@xxxxxxxxxxxx
> "Paul Andrews" <ac297@xxxxxxxxxxxxxxxxxx> wrote in message
> news:jprDe.3768$yH4.91@xxxxxxxxxxxxxxxxxxxxxxx
> > "Roy Hann" <specially@xxxxxxxxxxxxxxxxxxxxx> wrote in message
> > news:fKmdnXEEuftSpkPfRVnygg@xxxxxxxxxxxx
> > > > In an ideal world maybe they could press some button to get the
> desired
> > > > effect, but in practice there's more involved than pressing
buttons -
> > some
> > > > debate between humans!
> > >
> > > Exactly. That is precisely the point I am getting at. Humans need
> facts
> > so
> > > they can debate and decide. A job number is just an indirection to
the
> > > facts. Unless it adds value, it adds cost.
> >
> > The humans only need the job number to find the other facts. The use of
> the
> > job number simplifies lookup by humans and by query - reducing cost..
>
> I'm about to leave for sunnier climes so I am simply going to toss the
towel
> in here. One can discover examples where synthetic keys are truly useful,
> and I think we've agreed on that all along. My purpose in arguing about
> this has been to plant a seed of doubt that synthetic keys (and surrogate
> keys) are as useful as (I think) we generally assume. They're not.
>
> Roy

OK, have a great holiday!

Paul


.

Quantcast