IETF 120 - MEDIAMAN WG

Scribes:
Alexey Melnikov

Introduction, Note Well, scribe, agenda bash

Status of WG documents

• draft-ietf-mediaman-toplevel-06
• draft-ietf-mediaman-haptics-05

Haptics depends on top level draft, but both are in the RFC Editor's queue now.

• draft-ietf-mediaman-standards-tree

No feedback from attendees, chairs will start a WGLC on the mailing list.

• RFC6838 updates

Attendees at IETF 119 concurred there is a need to update 6838 and start work on 6838bis document, but the volunteer was not recorded in notes and there was not follow up. Chairs renewed request for volunteers and to move forward this work, potentially using the RFCXMLv2 content from the RFC Editors.

• [draft-ietf-mediaman-suffixes-08] (see next section)

Main topic for this session

Multiple suffixes for media types

Manu: As of -07 multiple suffixes (i.e. multiple plus signs) are now disallowed. No objections so far. Trying to clean up instructions for IANA and what is in the suffix registry.

Darrel: Speaking of the "leftmost plus sign" is weird if only 1 is allowed.

Manu: We can change this text.

Harald: I found some places where editorial changes might be needed. Can I please ask people to review the document before WGLC.

Mark Nottingham: Editor proposed a new draft, but it doesn't represent WG consensus. Some of the mailing list discussions are not reflected.

Mike Prorock: I have read it and did note one case of inconsistent language (not recommended and previous relevant requirement is a "MUST NOT"); we should clean that up.

Manu: We can fix that.

Manu: Mark, any specific changes that we missed?

Mark/Harald: Consensus by exhaustion?

Brian Campbell: Agreeing with Mark. Do we need this document at all?

Pete Resnick: No consensus for multiple suffixes. Doesn't seem to be a need for this in W3C now. But have we thought about eliminating suffixes altogether?

Manu: The original of this document: people didn't know what multiple suffixes mean, so we tried to clarify. So we need to publish something.

Manu: Registry cleanup work is needed anyway, this will need to be folded into rfc6838bis anyway.

Mike Jones: I think it is going too far not to allow suffixes entirely, as this will break existing RFCs (RFC 8725 and BCP 225).

Brent Zindel: W3C groups I am involved in no longer need multiple suffixes.

Murray: You don't want IESG to make decision on policy in this space.

Alexey: or Designated Experts make random decisions

Mark: Manu's proposal might be viable, but I am not sure that disallowing multiple suffixes now will prevent people asking for them in the future.

Manu: I talked to multiple W3C WGs and they agree not to register media types with multiple suffixes anymore. The only use case remaining is related to SVG?

Mark: Thank you, this looks like an impressive effort in convincing people in W3C not to use multiple suffixes.

Hans-Joerg: I might still have a use case for multiple suffixes for use in OSes, e.g. JSON-LD.

Manu: OS don't seem to be doing hierarchical fallback to generic handlers for XML/JSON/etc.

Manu: Possibly registering +JSON-LD instead?

Harald does a poll about disallowing multiple suffixes. Yes: 14 people, No: 1, No opinion: 3.

ACTION: Mark and Alexey to carefully review -08 to see if raised issues were addressed. Once they do, the WGLC can be requested.

Mark: is this going to be a standalone RFC or a part of -bis?

Harald: Both.

Harald: Reviewing milestones. I found one about Security Considerations checklist.

Murray reminded the WG that the milestone "Deliver any recommendations about future registration and queue management" is about managing the registry through GitHub idea.

Brian Campbell: What is the urgency in publishing the document about disallowing multiple suffixes? Nobody in W3C wants this anymore.

Harald: My personal opinion is that it is urgent to capture consensus of the community and finish the record of that consensus with a permanent record. I like finished, by IETF always leaves opportunity to re-open work if community changes.

Mark: Agreeing with Brian. Can have an agreement in the WG, but don't publish an RFC. Why can't we have the same input re urgency with the community registration process milestone for the WG?

Pete: We don't want to have another flood of requests with multiple suffixes in the future. We need to record our agreement somewhere. Possibly a Wiki, but getting an RFC done is better.

Harald: do we want to publish the multiple suffixes draft as an RFC: Yes: 7, No: 6, No opinions: 2.

Mark: my level of review will differ on whether the draft is to be published as an RFC or whether it will be incorporated into -bis.

Harald: please do a fairly thorough review.

Mark: I will do a "medium" level review ;-).

Rahul Gupta: RFC 2046 has no IANA registration template for any of media types. Is this an error or just the document is too old?

Darrel: IANA is fixing this, by creating a dummy registration files on the IANA website that say "this media type's registration entry occurred before later registration templates"