Re: Boeing XRI Use Cases

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Mark Baker writes:

> On 7/15/08, Henry S. Thompson <ht@inf.ed.ac.uk> wrote:
>>  Consider the ARK proposal (which I have always held up as a model of
>>  how to use http: URIs to address requirements similar to many of the
>>  requirements on XRI) [1].
>>
>>  It offers an approach in which e.g.
>>
>>       http://loc.gov/ark:/12025/654xz321
>>       http://rutgers.edu/ark:/12025/654xz321
>>
>>  identify the _same_ object. . .
>>
>>  Are you happy with that kind of design?
>
> No, for the same reasons mentioned by others here.

I want to dig a little deeper on this, to try to understand where the
bug is, in your opinion.

Suppose I and my friend start a club, and we agree that we will keep
duplicate copies of all the club records under the top-level 'club'
directory on our respective web-sites.  That is, we agree that
  http://www.thompson-example.org/club/xxx.html
and
  http://www.sabbatini-example.org/club/xxx.html
will identify the same resource and, to the best of our ability, that
we will respond with identical messages to GET requests to those
URIs.  Furthermore we make it a condition of joining our club that new
members do likewise.

Is this fundamentally at variance with Web Architecture?  I don't
think so.  AWWW does say [1]

   "Good practice: Avoiding URI aliases

    A URI owner SHOULD NOT associate arbitrarily different URIs with
    the same resource."

but in this case the two URIs are _not_ arbitrarily different, they
are in fact non-arbitrarily similar.

Of course such an arrangement has its downside: if a prospective
member of the club already _has_ a top-level 'club' directory, this
may appear to be a bar to joining the club, as they can't satisfy the
entry requirement stated above.  But there are obvious workarounds
(for instance, they ccould just add a new virtual host to their
server, called e.g. club.macavity-example.com, wrt which they can
serve
  http://www.club.macavity-example.org/club/xxx.html
etc. despite www.macavity-example.org/club already being in use)

Seems to me that

 a) This is a perfectly reasonable approach, which sensible folk might
    well adopt in order to protect against misunderstandings and/or
    lost data;

 b) It doesn't violate WebArch;

 c) It's actually exactly analogous to the ARK design.

If you've stayed with me this far, I'd be very grateful if you could
expand on where you disagree with the above analysis.

Thanks,

ht

[1] http://www.w3.org/TR/webarch/#uri-aliases
- -- 
       Henry S. Thompson, School of Informatics, University of Edinburgh
                         Half-time member of W3C Team
      10 Crichton Street, Edinburgh EH8 9AB, SCOTLAND -- (44) 131 650-4440
                Fax: (44) 131 650-4587, e-mail: ht@inf.ed.ac.uk
                       URL: http://www.ltg.ed.ac.uk/~ht/
[mail really from me _always_ has this .sig -- mail without it is forged spam]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (GNU/Linux)

iD8DBQFIfmFCkjnJixAXWBoRArQtAJwKmuSnWPwu9TuTVZtrTakwrgvDlgCfezbz
Cx5nvsIIS1fcavOJ8IwykX8=
=1YRp
-----END PGP SIGNATURE-----

Received on Wednesday, 16 July 2008 21:00:38 UTC