Re: Comments solicited: "Providing and discovering definitions of URIs"

Hi Jonathan,

On Sat, 2011-06-25 at 16:12 +0000, Jonathan Rees wrote:
> Comments solicited: "Providing and discovering definitions of URIs"
>   http://www.w3.org/2001/tag/awwsw/issue57/20110625/
> (this is TAG ISSUE-57 / ACTION-579)

Some comments:

1. Regarding Section 1.1 Success criteria #6: "Compatible with Web
architecture. A URI should have a single agreed meaning globally,
whether it's used as a protocol element, hyperlink, or name."  To avoid
the issue of what is meant by "meaning" (which *is* a real issue, as
there is a big difference, for example, between assuming that a URI has
a globally unique interpretation and assuming that its interpretations
are merely globally *constrained* by its definition), I suggest changing
this criterion to: "Compatible with Web architecture. A URI should have
a single agreed definition globally, whether it's used as a protocol
element, hyperlink, or name."


2. It would be helpful in Section 3 if each method of definition were
demonstrated in terms of the use case given in Sec 2.1, saying exactly
what Alice, Bob and Carol do when that method of definition is used.


3. Another criticism of the 4.2 approach (though debatable) should be
that indirection is unnatural for users who wish to make statements
about the movie: the 4.2 approach forces them to think indirectly.


4. It would be helpful to expand the sec 2.1 use case to include the
situation that will cause the ambiguity issues that are later discussed
to arise.  This would allow this same use case to be applied uniformly
to all potential solutions.  For example, you might add a character,
Derek, who writes a document containing metadata about Alice's
definitional document; a character, Erin, who publishes a document
containing the merge of Bob's document and Derek's document; and a
character Frank who reads Erin's document and needs to avoid confusing
statements about the earthquake with statements about Alice's
definitional document.  You may need to change the use case a little to
make the ambiguity more apparent -- so that Bob and Derek use the same
property.



5. Typos:

"Location: header specifying 'http://example/eq018' as the redirect
target" should be: "Location: header specifying
'http://example/about-eq018' as the redirect target"

s/secion/section/

s/variety contexts/variety of contexts/

s/probably better use/probably better to use/

s/might eliminate/might reduce/


-- 
David Booth, Ph.D.
http://dbooth.org/

Opinions expressed herein are those of the author and do not necessarily
reflect those of his employer.

Received on Thursday, 30 June 2011 18:50:40 UTC