Re: Sets and Lists, again

JOG wrote:
> In the end a list is 'this, then that, then t'other' - purely ordinal.
> As such here I agree with Gene - creating an index to identify the
> location of each item in that list is a physical accessor.

I sure don't see that at all. How is it any more physical than any other name to call something emailAddress[2]? It is logically the 3rd (or 2nd) element in a list, but it need not be a physical designation at all. This is definitely a logical designation, with no prescription for anything physical (although it could obviously relate to a physical design).

> To obtain
> the third element in a list, for example, one processes 'start_head,
> get_next, get_next'.

Now you are talking physical matters. I'm speaking purely at the logical level when working with a list-valued attribute.

> Any cardinal access to 'element[3]' is surely a
> physical, and not logical, shortcut.

I really am not following your logic on this. It sounds like your argument amounts to saying that the type of "list-valued attribute" is necessarily a physical and not logical type. That makes no sense, does it? --dawn Received on Tue May 23 2006 - 04:28:14 CEST