IRI WG IETF 82, Taipei Materials: Agenda: http://www.ietf.org/proceedings/82/agenda/iri.txt WG Overview: http://www.ietf.org/proceedings/82/slides/iri-2.pdf IRI 3987bis Issues Overview: http://www.ietf.org/proceedings/82/slides/iri-3.pdf IRI 4395bis Issues: http://www.ietf.org/proceedings/82/slides/iri-0.pdf Agenda: 1. Admin (scribes, co-chairs, ...) 2. Report of W3C discussions 3. Guidelines and Registration Procedures for New URI/IRI Schemes draft-ietf-iri-4395bis-irireg 4. Internationalized Resource Identifiers (IRIs) draft-ietf-iri-3987bis 5. Internationalized Resource Identifiers (IRIs) Comparison draft-ietf-iri-comparison 6. Internationalized Resource Identifiers (IRIs) BIDI draft-ietf-iri-bidi-guidelines 1. WG Status: - Since IETF 81 4395bis has been nearing last call status, but still has a dependency on 3987bis -- 4395bis issues have been addressed and many are ready to be closed - Since IETF 81 the decisions were to split up 3987bis into several documents: -- One describing IRIs generally -- One describing IRI comparison -- One describing IRI BIDI rules -- A decision was also made to move processing material out to a separate "processing spec" which has since not been realized. -- The "processing spec" has since been considered best moved to the W3C to describe URI and IRI parsing and processing - Today's meeting will be an open discussion on each of these documents, moderated by Peter 2. W3C Coordination / URI/IRI Processing Spec - processing spec done by W3C. would be a self-contained document that IETF could reference. 3. Guidelines and Registration Procedures for New URI/IRI Schemes draft-ietf-iri-4395bis-irireg - there is work going on to make registration processes easier in general ("happpiana" discussions) - Larry notes that there are two aspects, (1) IRI-enablement of the URI scheme registry and (2) making registration easier - Larry suggests finishing the i18n-related (non-registry process related) work and then working on the registration process topics in coordination with/after happyIANA process converge. - #70: instead, text to guide the expert review. - #79: have columns for submitter, schema "owner", but rely on expert reviewer to agree on the change. IANA has similar for ports registry (registrant, contact). - #97: would do a change of state. Need to define the list of states for each entry. - #98: close the ticket. will be handled in the IANA considerations section when we are done. - #73 and #106 be unified (Martin didn't see #73 when he created #106) - #83: disagree. close the ticket - new ticket? need to provide better guidance to registrant and expert with regard to security considerations (which *might* be "unknown, use with care") - not discussing the minor tickets. moving on to next agenda item. WG participants encouraged to review these issues and post to the list. 4. Internationalized Resource Identifiers (IRIs) draft-ietf-iri-3987bis #105: to be discussed between co-chairs and AD about intended target status of document #3: split, assign to relevant 'component', address individually (many might no longer be relevant) #85: accept in principle: align with terminology in RFC 6365. todo for editors. review by 6365 authors or WG participants might be helpful. #5: no resolution. Larry: we have a way to detect uri in text. we don't have text for detecting iri in text. should this be a new document? ticket to be opened. JKlensin: out of scope for the wg charter. #13: May be out of data, because it refers to IDNA 2003. Larry to check/update to allign with IDNA 2008. #14: Larry to split text and send to the list, since the whole ticket is complicated to review. #92: editorial. skipping #93,103,104: Martin would do the first: move back to LEIRI. silence. no oppose. to the editor to do. Section 6, Characters Disallowed or Not Recommended in IRIs, was a subsection of LEIRI, carefully negotiated with W3C XML Core WG #89: Peter: is this scheme-specific? Martin: for some, yes. #91: John: Danger of starting with something in NFD. Pete: Clarifying question: Are you saying that for NETBios or Wins, it will first resolve, and then use HTTP? Reply: Yes? Pete: Next question: Do we have something that can take a %-encoded domain name and automatically un-escapes it. Rely: We don't know. no resolution.