Re: Comments on Section 3

On Mon, 14 Apr 2003, Kevin Smathers wrote:

> The main problem that I have with this approach is that I don't consider
> two files to be the same just because they happen to have the same
> content.  Likewise two facts. ...  Imagine how you would implement
> default constructors, or access controls.

What about two integers? I guess I view a file instance as a distinct from
its contents.  If you have two copies of a book then some metadata is
attached to the book itself, such as the author and title, and other
metadata is attached to the copies, such as loan status or aquisition
date.

>> The scenario I envision here would be that of a photographer who wanted
>> to catalog photos of downtown Boston and needs to find a robust name
>> for "Boston".

> This is the problem of vocabularies, listed elsewhere in the use case
> document.

Names are slightly different than controlled vocabularies in that
controlled vocabularies are a fixed set of names while a naming authority
can produce an unbounded set of names.  But even this is a weak
distinction because a controlled vocabulary can always be revised to
include new terms or names. I agree that solutions for discovery of
controlled vocabularies will work just as well for discovery of naming
authorities.

Nick Matsakis

Received on Monday, 14 April 2003 18:34:58 UTC