There was earlier some discussion about whether to allow IRIs (instead and on top of URIs) in Atom. The decision was put off because some people had doubts whether IRIs would move ahead on the standards track or not, and if they'd move on, how fast.
As you may already have seen, IRIs have been approved as an IETF Proposed Standard last week (see also below). I therefore propose that we use IRIs in Atom. Unless there is strong objection already at this point, I'll go ahead and work out the pace that I already started (http://intertwingly.net/wiki/pie/PaceIRI).
Regards, Martin.
>From: The IESG <[EMAIL PROTECTED]>
>To: IETF-Announce <[EMAIL PROTECTED]>
>Date: Wed, 08 Dec 2004 09:55:16 -0500
>Cc: Internet Architecture Board <[EMAIL PROTECTED]>,RFC Editor <[EMAIL PROTECTED]>
>Subject: Protocol Action: 'Internationalized Resource Identifiers (IRIs)' to Proposed Standard
>The IESG has approved the following document:
>
>- 'Internationalized Resource Identifiers (IRIs) '
> <draft-duerst-iri-11.txt> as a Proposed Standard
>
>This document has been reviewed in the IETF but is not the product of an
>IETF Working Group.
>
>The IESG contact person is Ted Hardie.
>
>Technical Summary
>
>This document describes Internationalized Resource Identifiers and their
>relationship to URIs. While the character limitations of URIs are not
>usually an issue for protocol processing, they may restrict the usefulness of
>the identifiers presented to end users or systems expecting a different range of
>characters. Rather than extend URIs, this document introduces a new identifier
>type and a describes a relationship to URIs. Within an IETF context, IRIs
>will likely be used as presentation elements. There are cases, such as XML
>namespaces, in which an IRI may be used as a token, because
>character-by-character equivalence is the only property used for protocol
>processing. In no case should an implementor assume that an IRI may be
>substituted for a URI in an existing protocol grammar; either the generative
>grammar associated with the protocol must be updated to specify IRIs or the
>implementation must transform an IRI into a URI before use.
>
>Working Group Summary
>
>This work was initiated in the W3C, and it has been broadly accepted in that
>context. It has also been discussed on the URI mailing list and a public, open
>list ([EMAIL PROTECTED]) dedicated to the topic. Considerable care has been
>taken to keep this specification well-synchronized with the URI specification.
>There were issues raised during IETF Last Call, and a new document version
>resolving those issues was submitted.
>
>
>_______________________________________________
>IETF-Announce mailing list
>[EMAIL PROTECTED]
>https://www1.ietf.org/mailman/listinfo/ietf-announce
