Re: Idempotence and "Replication Insensitivity" are equivalent ?

Marshall ha scritto:

> pamelafluente_at_libero.it wrote:
> > Marshall ha scritto:
> > > >
> >
> > Hi Marshall :) , Hmmm,
>
> Buongiorno Pamela :-) Or as we say in the USA: "hey."
>
>
> > T T F should evaluate to F
> > T would be in contrast with NOR definition:
> >
> >
> > "A predicate in logic equivalent to the composition
> > NOT OR that yields false (F) if ANY condition is true,
> > and true (T) if ALL conditions are false. "
>
> You're still not getting the relationship between aggregate
> functions and binary functions. Until you understand this
> relationship this whole area is going to confuse you.
>
> aggregate-NOR:
> "duplication-sensitive"
> defined as fold(NOT(OR)) = fold(NOR)
> NOR: binary, not idempotent
>
>
> n-ary NOR:
> not "duplication sensitive"
> defined as NOT(fold(OR))
> OR: binary, idempotent
>
> Note again the relationship between idempotence of the function
> being folded and the "duplication sensitivity" of the resulting
> aggregate.
>
> Uh, ciao, I think it is.
>
Right Ciao!

Well if NOR of ( T T F ) to you evaluates to T, while the definition says that it should evaluate to F, I am totally confused.

Further, Chris, who you think highly of, seems to have a different opinion, and he is saying the counterexample works.

Hmmm...

Well actually I have a different way to deal with math. I usually visualize intuitively the result first, then try to formalize it, just to make sure intuition is right.

Working with formalism only can be dangerous (to me), as it's kind of semi-blind walking. Clearly formalism is essential to make sure that we got it right. I do not know if professional mathematicians here follows this way in their higly creative work or proceed with formalism only ...

To my intuition I think I got it right, but have problems to follow you in that kind of formalism that, frankly, to me seems an unnecessary complication for concepts that are quite simple. Further in practice, I feel, it's not much useful, as we program and compute all sort of aggregate functions (KPIs, etc.).

I leave to you and Chris the last discussion on this very matter. To me idempotence and duplication insensitivity are not the same thing at all (at least intuitively).

We have done a little battle on that (didn't we ? :) ) and after all we have improved our understanding of these things.

In case I'd like to return to the original problem of subquery construction and query splitting. I am just a programmer, remember ? ...

-P Received on Tue Sep 19 2006 - 22:32:32 CEST