Re: Impossible Database Design?
Date: Mon, 7 Oct 2013 00:53:39 -0700 (PDT)
Message-ID: <0b473ffa-3fe4-4e41-99a8-e748af4671c8_at_googlegroups.com>
On Monday, October 7, 2013 1:05:08 PM UTC+8, Jan Hidders wrote:
> > I believe in the context of the somewhat superficial discussion (about
> > finite versus infinite), the statement was intended to suggest that
> > infinite data types are not useful to computer science. If so I'd call
> > that a gross misrepresentation!
>
> Me too, but did it really?
It appears so. In another post Bob said:
If one accepts the limitation of a finite representation, which in the end one has no choice over, then there will always be a finite endpoint at the extreme end of the representation.
which is incorrect.
> I assume by infinite type we mean here a type
> with infinitely many instances.
Yes
> While clearly representable in a database,
> it makes the life of the DBMS easier if the size of the representation is
> more predictable.
Yes and the proper use of finite types is certainly very important for efficiency.
However I wouldn't say finite types are necessary or sufficient for predictability in general. E.g. it's all too easy to define an innocuous looking finite relation type which allows more tuples than the number of atoms in the observable universe.
David Received on Mon Oct 07 2013 - 09:53:39 CEST