RE: [XRI] XRI-as-Relative-URI proposal (ACTION-189 refers)

Henry,

Thanks very much for the feedback. Sorry to be slow in responding -- last
week we held the XRI TC F2F meeting after Internet Identity Workshop in
Mountain View.

Thankfully you posted your message before the meeting so we were able to
discuss it there. See responses inline.

> -----Original Message-----
> From: Henry S. Thompson [mailto:ht@inf.ed.ac.uk]
> Sent: Thursday, November 13, 2008 3:41 AM
> To: Drummond Reed
> Cc: www-tag@w3.org; 'Peter Davis'; jbradley@mac.com
> Subject: Re: [XRI] XRI-as-Relative-URI proposal (ACTION-189 refers)
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Drummond Reed writes:
> 
> > [There's] a new proposal for how XRIs can better fit with AWWW
> > architecture.
> > . . .
> > The proposal is written up on an XRI TC wiki page at:
> >
> > 	http://wiki.oasis-open.org/xri/XriAsRelativeUri
> 
> 
> I've looked at this in some detail, and agree with others' comments
> that it looks like a very good direction to move in.
> 
> I think the clarifications that have emerged in this thread wrt what
> you call the concrete/abstract distinction are important, and should
> be included, very carefully, in the eventual full specification.
> 
> I have two remaining concerns:
> 
>  1) Very little is said about where the necessary base URIs are going
>     to come from.  This is almost certainly corrigible going forward,
>     that is, in a new specification, although it will have to be done
>     carefully, and, I hope, with reference to either the Infoset [base
>     URI] property [1] or XML Base [2] insofar as XML-expressed
>     languages are concerned.  Retrospective cleanup, for example wrt
>     OpenID's usage of 'short' old-style XRIs such as =jbradley, will
>     be more challenging, but should none-the-less be attempted I
>     think.

We agree that the XRI 3.0 specification must provide clear guidance about
the base URI. We will be proactive about asking the TAG for review of the
XRI 3.0 syntax and bindings specs as they proceed.

>  2) I think the list of candidate schemes which may be used in
>     XRI-signalling base URIs is too ambitious, unnecessarily so as far
>     as I can see.  Including e.g. ftp: is surely to miss the 80-20
>     point by a long way.  More seriously, including urn: is a big
>     mistake.  There is no well-defined notion of relative URN, or of
>     absolutisation for URNs.  URNs are not what RFC 3986 calls
>     'hierarchical identifiers', and the RFC says explicitly [3]:
> 
>        "relative references can only be used within the context of a
>         hierarchical URI"
> 
>     Please just don't go there!

The XRI TC agrees with you that XRI bindings should be to hierarchical URI
schemes. At the F2F meeting we concluded that the initial three XRI bindings
will be to http:, https:, and info:. We will consider additional bindings as
required over time, but these three should give us plenty to get started
with.

Best,

=Drummond 


> [1] http://www.w3.org/TR/xml-infoset/#infoitem.element
> [2] http://www.w3.org/TR/xmlbase/
> [3] http://www.apps.ietf.org/rfc/rfc3986.html#sec-1.2.3
> - --
>        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) 131 651-1426, 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)
> 
> iD8DBQFJHBJIkjnJixAXWBoRAlZeAKCA92xk8cKVybPpNW3oNj2JlLiJ1gCfehZf
> DKtmOZjTB3RSLDfEktpUDoI=
> =rif+
> -----END PGP SIGNATURE-----

Received on Tuesday, 18 November 2008 19:42:31 UTC