Re: Table design question

"Bob Badour" <bbadour_at_golden.net> wrote in message news:iqmdnQhOysUkBYfdRVn-jg_at_golden.net...
> "Mike Sherrill" <MSherrillnonono_at_compuserve.com> wrote in message
> news:qimk1018kh3htkap6gkdif05j1q8705v6v_at_4ax.com...
> > On 25 Jan 2004 03:39:33 -0800, andrewst_at_onetel.net.uk (Tony) wrote:
> >
> > >But it isn't a domain: the domain of user ID numbers would be
> > >something like "any 6 digit number between 100000 and 999999" or "a
> > >string of between 3 and 30 letters and numbers"; whereas this table
> > >represents the list of users who actually exist,
> >
> > Try thinking about it this way . . .
> >
> > At the conceptual level, a domain is just a data type, and a type is,
> > among other things, a set of all possible values. One way to handle a
> > set of values is to store them in a table.
>
> It's the "among other things" that kills you. A variable is not a time
> invariant set of values and their associated operations.
>
>
I'm not sure I understand, Bob.

A variable holds a single value drawn from some domain. By virtue of the fact that the domain has been defined at some point in time as a set of all possible values with operators, it should be time-invariant as long as the type specification holds. Why wouldn't it be invariant? Dynamically changing the specification (and therefore its definition) of some variable in terms of its type means redefining both type and domain as something different - a new type.

Computing devices are constricted to rather finite sets of representations, as is logic. Are you saying that this constraint actually does not exist?

• Dan