Re: Domain

Bob Badour <bbadour_at_golden.net> schrieb:
> As long as the dbms treats values strictly as values at the logical
> level, all values are atomic. A dbms would violate atomicity if it
> exposes internal structure with concepts such as a current position
> within the value, for instance.

I don't think it's so clear-cut. What's the difference between a function "value -> current position within value" and a function "date -> month"? What if cons cells were used for all non-scalar types. Would car and cdr "expose internal structure"?

I'd like to say, if A is a subtype of B (by which I mean, the set of values and operations on B are subsets of those of A), then A is non-atomic, since it can clearly be divided into B and non-B parts. I think this matches the intuitive sense of the word pretty well. Received on Sat Jan 03 2004 - 20:47:59 CET