Minutes of the 2021-01-13 IAB Teleconference
- Jari Arkko
- Ben Campbell
- Alissa Cooper (IETF Chair)
- Stephen Farrell
- Wes Hardaker
- Cullen Jennings
- Mirja Kühlewind (IAB Chair)
- Zhenbin Li
- Jared Mauch
- Cindy Morgan (IAB Executive Administrative Manager)
- Karen O’Donoghue (ISOC Liaison)
- Tommy Pauly
- Colin Perkins (IRTF Chair)
- Alvaro Retana (IESG Liaison)
- Jeff Tantsura
- Amy Vezza (IETF Secretariat)
- Jiankang Yao
- Mark Nottingham
- Warren Kumari
- Greg Wood
1.2. Agenda bash & announcements
Cindy Morgan reminded the IAB that the BOF Coordination call for IETF 110 is on 2021-01-14 at 2130 UTC.
1.3. Meeting Minutes
The following meeting minutes remain under review:
- 2021-01-06 business meeting – (draft submitted 2021-01-06)
1.4. Action item review
- 2021-01-06: Cullen Jennings to start an email thread with Alissa Cooper, Mark Nottingham, and Martin Thomson about the W3C and the completion of the WebRTC spec.
- 2021-01-06: Mirja Kühlewind to start a discussion on the IAB list about active involvement in discussions that relate to broader Internet changes and IAB support for new work.
- 2021-01-06: Wes Hardaker to reply to Ben Kaduk’s email about the DANE for IoT onboarding BoF proposal.
- 2021-01-06: Wes Hardaker to schedule a call with the IAB and the DANE for IoT onboarding BoF proponents.
- 2020-10-14: IAB to review the Temporary RFC Series Project Manager contract / RFC Editor future development process. January 2021.
- 2021-01-06: Jared Mauch to organize a side meeting on DDoS/attack protection. Due April 2021.
- 2021-01-06: Stephen Farrell, Ben Campbell, Cullen Jennings, Mirja Kühlewind, and Jared Mauch to interview the ISOC BOT candidates.
- 2021-01-06: Cindy Morgan to schedule the ISOC BOT interviews.
- 2021-01-13: Wes Hardaker with help from other IAB members to add characteristics of readable RFCs (with examples) to https://www.iab.org/wiki/index.php/RFC_Readability. (Deadline: 2021-02-17)
2. Cluster 238 and Complex Inter-Spec Dependencies
Cullen Jennings briefly described how the WebRTC work led to Cluster 238. The 1500-2000 pages of new IETF specifications spanned two SDOs, three IETF Areas, and many Working Groups. Those specifications normatively reference around 5000 pages of WebRTC specific specifications.
Cullen Jennings suggested that the WebRTC interdependencies are so complicated because:
- WebRTC profiles SIP and removes stuff not needed
- WebRTC mandates a minimum system implementation that ensures interopability
- WebRTC is backwards-compatible with most current SIP deployments
He added that everyone in the IETF has things that they think are important, and the easiest way to get consensus is to add all of those things into the spec. Because the process for updating a published RFC is so arduous, everyone wants to get everything included in the first version because there may not be a second version.
Ben Campbell noted that since WebRTC was being built on top of existing technologies, there were many existing working groups that were each in charge of part of the system.
Alissa Cooper said that while she understands why documents are held over normative references, she is not sure whether all of the normative references in the WebRTC documents are necessary.
Jared Mauch said that the situation could have been better if the IETF had living documents where references could be updated after publication.
Tommy Pauly asked at what point it became clear that there would be so many normative dependencies. Cullen Jennings replied that it was clear fairly early in the process. Tommy asked if they could have untangled the dependencies before they got entrenched. Cullen replied that they may have been able to, but the people working on WebRTC were more focused on the content of the spec than the organization of it.
Alissa Cooper suggested that the problem is not the number of dependencies, but that the RFC publication process makes it so difficult to update a published RFC. If the references in an RFC could be more easily updated, the number of normative dependencies wouldn’t be a barrier for publication.
Mirja Kühlewind asked whether there was anything learned during this process that the IAB should document.
Ben Campbell said that most IETF RFCs are written for an IETF audience, but many people who need to implement IETF documents are not in the IETF community. Wes Hardaker agreed, noting that most other publishers focus on their readers in a way that the IETF does not.
Cullen Jennings mused that it would be nice if there was something like an “RFCs for Readers” requirements document.
Wes Hardaker suggested that the IAB could write something that outlines the properties of readable documents and cites examples.
Jared Mauch said that that would make an interesting blog post.
Wes Hardaker agreed to set up a page on the wiki to brainstorm about the characteristics of readable RFCs (see https://www.iab.org/wiki/index.php/RFC_Readability).