Re: draft-masinter-dated-uri-00.txt

On Sunday, August 26, 2001, at 02:51  AM, Larry Masinter - 
LMM@acm.org wrote:

> # In lieu of
> # <urn:duri:19990114:http://www.w3.org/TR/REC-xml-names> you may
> # want to cite <http://www.w3.org/TR/1999/REC-xml-names-19990114>.
> # Can you explain why you didn't?
>
> Well, the latter depends on the good graces of the webmaster
> of www.w3.org to not fool around and decide that, for example,
> "1999" should come before "TR" instead of after.

Ehm, I don't see how that is true. A problem only would arise if 
the webmaster decided to use this URI to identify something 
else, which seems highly unlikely. If the webmaster simply 
decided that the 1999 should come first, both URIs would work 
equally well -- that is they would clearly identify a resource 
but that resource would not be easily accessible by say, typing 
the URI into a web browser.

> It's one thing to urge webmasters to not move content around,
> and it's quite another to build systems that break if they don't
> listen and do.

I don't see how your citation would be any more fragile either way.

> # Using:
> #         urn:duri:2000:urn:ietf:std:50
> # to refer to "the document that was STD 50 that was in effect as
> # of the first instant of 2000" seems problematic to me, since
> # that URN represented the same resource in 2000 as it would at
> # any other time, namely the IETF STD 50 document. You seem to be
> # implying that the use of duri changes the semantics of the
> # resource involved. Is this so?
>
> The IETF explicitly reassigns the numbers in the STD series as
> documents get updated, and doesn't reassign RFC numbers. The
> use of "duri" doesn't change the semantics of the resource, just
> of the reference. It fixes the reference in time. At the beginning
> of 2000, "urn:ietf:std:50" referred to RFC 1643.

No, it identifies a Resource (STD 50), a conceptual mapping 
which corresponds do a different "entity" over time.

> If the RFC gets updated, STD 50 will be a different document, but
> urn:duri:2000:urn:ietf:std:50 will refer to the same document
> it did in 2000.

So you're saying that using duri does a bit of "resource 
resolution" -- that a duri URI points not to a resource at a 
certain point of time, but instead the entity that that resource 
corresponded to? This is an important distinction.

--
[ "Aaron Swartz" ; <mailto:me@aaronsw.com> ; <http://www.aaronsw.com/> ]

Received on Sunday, 21 October 2001 20:46:38 UTC