Re: Data Display & Modeling

"mAsterdam" <mAsterdam_at_vrijdag.org> wrote in message news:409d581d$0$559$e4fe514c_at_news.xs4all.nl...
> Hm. How about this: 'atom' for the the atomic attributes
> (hesitation: possible prolog clash), 'compound' for
> the structered attributes (maybe just 'list' for list),
> and 'chunk' for the construct which *only* differs
> from a RM.relation in that it's attributes maybe compound.

These distinctions are not really distinctions. Is a string compound? If not, why not? How is a list value different from a date value? I can pluck the month and year from a date, can't I? Both of them can hide HOW the data is stored internally (e.g. a date could just be a string, while a list can be a linked list).

If we can't define "compound" (and we can't!), we can't define "chunk" as something other than a tuple (which I assume you mean rather than a relation).

> Reiterating:
> - relations have atoms only as attributes

Say "values", as distinct from "variables".

> - chunks have compounds as well as atoms as attributes

See above.

> - a list is a special kind of compound

In what way is it special?

> - model R only allows relations
> - model H allows relations as well as chunks (this is the only
> difference between H and R).

H == R, though R discourages what H explicitly enables, in favor of further decomposition.

> Then, to prevent nasty gotchas:
> Let's keep the distinction clear.
> Atoms and compounds.

No distinction, at least none that's definable or useful.

• erk