Minutes IETF 109 BOF coordination interim-2020-iesg-23 2020-10-09 14:00
minutes-interim-2020-iesg-23-202010091400-00

IETF 109 BoF Coordination Call
9 October 2020

Reported by: Amy Vezza, IETF Secretariat
Revised with notes from:

Additional reference materials available at the BoF Wiki
(https://trac.tools.ietf.org/bof/trac/).

ATTENDEES
---------------------

Jari Arkko (IAB)
Deborah Brungard (RTG)
Ben Campbell (IAB)
Alissa Cooper (GEN)
Roman Danyliw (SEC)
Martin Duke (TSV)
Stephen Farrell (IAB)
Liz Flynn (Secretariat)
Wes Hardaker (IAB)
Cullen Jennings (IAB)
Benjamin Kaduk (SEC)
Erik Kline (INT)
Mirja Kühlewind (IAB Chair)
Warren Kumari (OPS/MGT)
Barry Leiba (ART)
Zhenbin Li (IAB)
Jared Mauch (IAB)
Cindy Morgan (Secretariat)
Karen O'Donoghue (ISOC)
Tommy Pauly (IAB)
Colin Perkins (IRTF)
Alvaro Retana (RTG)
Amy Vezza (Secretariat)
Martin Vigoureux (RTG)
Éric Vyncke (INT)
Robert Wilton (OPS/MGT)
Jiankang Yao (IAB)

REGRETS
------------------------

Murray Kucharawy (ART)
Mark Nottingham (IAB)
Jeff Tantsura (IAB)
Magnus Westerlund (TSV)

APPLICATIONS AND REAL-TIME AREA

o A Semantic Definition Format for Data and Interactions of Things 
(ASDF)
Responsible Area Director: Barry Leiba

The proposed Working Group A Semantic Definition Format for Data and 
Interactions of Things (ASDF) was added to the BoF Wiki to make sure 
the group was able to get a session to meet at IETF 109. The charter 
was approved pending edits, and it is expected to be a working group 
by IETF 109.

o Building Blocks for HTTP APIs (HTTPAPI)
Responsible Area Director: Barry Leiba

The proposed Working Group Building Blocks for HTTP APIs (HTTPAPI) was
added to the BoF Wiki to make sure the group was able to get a session
to meet at IETF 109. The charter is in process, and it is expected to
be a working group by IETF 109.

o JSON Path (JSONPATH)
Responsible Area Director: Murray Kucharawy

The proposed Working Group JSON Path (JSONPATH) was added to the BoF 
Wiki to make sure the group was able to get a session to meet at IETF 
109. The charter is in process, and it is expected to be a working 
group by IETF 109.

GENERAL AREA

NONE

INTERNET AREA

o MAC Address Device Identification for Network and Application 
Services (MADINAS)
Responsible rea Director: Éric Vyncke

Éric Vyncke introduced the MADINAS BoF effort. He said he thought it 
would be worth having the BoF, but he wasn't sure how the work would 
evolve as it could intersect with IEEE and DHC, and that it was also 
semi-related to CAPPORT. He also noted that the proponents were mostly 
all from one company. Éric mentioned he has two possible chairs, but 
would like additional choices to diversify point of view. 

Warren Kumari suggested that a chair who works in privacy might help 
diversify.

Alissa Cooper noted she couldn't tell from the description what the 
main idea of the BoF was: developing APIs or recapturing stability of 
identifiers.

Éric agreed but also noted that the community should be the ones 
making that decision via the BoF discussion.

Cullen Jennings mentioned that the wiki and the discussion do not seem 
to match what work the BoF would work on.

Alissa noted that could create a contentious debate.

Barry Leiba disagreed with Cullen's assessment, and said he believed 
the BoF would figure out a way to make sure applications and services 
worked correctly even as MAC addresses change.

Stephen Farrell noted that he asked a question via email that he did 
not receive a reply to from the proponents. He believed that without 
narrowing the scope of the BoF, the discussion may generate 
divisiveness.  He noted that the proponents should decide how they 
expect to handle privacy before they have the discussion.

There was a short discussion on privacy and getting services running 
on the network level. The examples stated were about the necessity of 
re-authenticating and also the wire messaging service wiping out a 
chat history each time a MAC address changed.

Éric noted that the BoF description could be improved to clarify the 
focus of the BoF, but he believed it to be a hot topic. 

Stephen agreed the BoF should be run, but wanted to see a revised 
description before IETF 109. He expressed concern over the positioning 
as written in the BoF description.

Jari Arkko thought reframing the BoF description could help the BoF 
generate a more constructive discussion.

The IESG and IAB had a short discussion on the BoF process and how to 
work with the proponents to revise the description of the BoF. 

Tommy Pauly noted that choice of chairs could help focus and narrow 
the BoF topic.

Éric agreed that strong chairs would be crucial to the success of the 
BoF.

The MAC Address Device Identification for Network and Application 
Services BoF was provisionally approved for IETF 109. Éric Vyncke with 
help from Jari Arkko and Stephen Farrell will work with the BoF 
proponents on framing the discussion and to narrow the scope of the 
proposed BoF.

OPERATIONS AND MANAGEMENT AREA

NONE

ROUTING AREA

o Application-aware Networking (apn)
Responsible Area Director: Martin Vigoureux

Zhenbin Li, who participated in this meeting as an IAB member, is one 
of the proponents of this BoF. This prompted a short discussion on the 
BoF process on having a BoF proponent present for the conversation.

Martin Vigoureux introduced the proposed APN BoF and noted that the 
last time this BoF was proposed it was declined.

Ben Kaduk asked about the side meeting the APN proponents ran the last 
time the BoF was proposed. 

Zhenbin Li mentioned he had interest in the work and was happy to 
answer any questions about the proposed BoF.

Martin V. said that the refusal from the last meeting was more because 
the proposal was not clear enough, but since then, the proponents have 
worked to refine the topic for the BoF and set up a mailing list to 
discuss the issues. He had discussed the possibility of another side 
meeting, but the proponents mentioned to him they need more people to 
be interested in the topic for constructive discussion. The proponents 
are requesting a non-wg forming BoF to gauge interest.

Barry Leiba noted that the last time the BoF was proposed, he had been 
concerned about where in the stack the proposal would focus their 
work. Since the revision of the proposal pushes the work to the lower 
stack layers, he thinks it is better focused and it would be a good 
time to have a BoF to discuss it.

Tommy Pauly agreed a non-WG forming BoF makes sense. He also mentioned 
that the description needs refining, and the claims about security and 
privacy were too strong. He thinks that with a little bit of editing 
of the description the BoF could be successful.

Martin V. agreed.

Tommy noted that the name of the BoF includes "application-aware," but 
that there is a difference between something like a network 
application and what a user would think of as an application on their 
phone. He asked which definition of "application" was meant here: is 
this about what the user thinks of as an app, or is it about the more 
abstract protocol?

Zhenbin replied that this is focusing on the network layer and limited 
domains; SRv6 is the example used in the BoF proposal. He said that he 
thinks it is not mandatory for the user to explicitly specify the 
application in a limited domain if the information can be gleaned from 
the port number or interface that is connected. VLAN tagging can be 
used to identify the user and the service. It would be a universal 
identification but the application information would be out of scope 
because it is too complex.

Colin Perkins noted that the IETF has had other BoFs that were 
similar, and the PANRG is currently doing some research in similar 
spaces. He suggested the proponents could talk to the PANRG about it, 
as the boundary between research and engineering was unclear.

Warren Kumari stated that the description seemed like it was looking 
to build a framework that could be used for other applications. He 
suggested it looked similar to network tokens and said he was worried 
about user privacy on the limited domain space. He also suggested that 
if the research group was doing similar things, the discussion could 
be had there.

Martin V. agreed that presenting in PANRG may be a way forward, 
however he was still inclined to let the BoF happen to get more views. 
He added that it wasn't a BoF to do the work, but a BoF to discuss and 
define what work can be done in that space.

Warren mentioned that network slicing was a similarly vague 
description and was very contentious as a topic.

Martin V. replied that a lot of working groups are now working on 
various aspects of network slicing. He noted that while this 
particular BoF effort may not become a working group, other groups may 
pick up various parts of the work.

Alissa Cooper observed that the proponents all seem to be from one 
side of the equation. She mentioned she would like to see application 
providers be part of the discussion. She also wondered whether the 
Routing Area was the right place for the discussion. She noted that 
the Transport Area would likely have a lot of interested parties.

Jari Arkko agreed that application providers should be part of the 
discussion. He also noted that a change in perspective might help the 
proponents of the work refine the description by reframing the work in 
terms of the service provider perspective.

Zhenbin noted that during a previously held side meeting, the group 
clarified that the work would be done in the network layer rather than 
using the network token technology in the transport layer.

Ben Kaduk mentioned he was uncomfortable with the statements in the 
description about the lack of security concerns; however, he didn't 
want to block the discussion.

Mirja Kühlewind said that she wanted to see a more concrete proposal 
to narrow the work. She was concerned there was no clear unsolved 
problem to be discussed.

Tommy agreed that the BoF proposal could be narrowed in scope.

The IESG and IAB had a short discussion on the BoF process.

Éric Vyncke noted that it was normal for BoFs to sometimes be unclear 
and the discussion generally helps narrow the scope of the work. He 
also said he'd rather see the work discussed at the IETF.

The Application-aware Networking BoF was provisionally approved for 
IETF 109. Martin Vigoureux will work with the BoF proponents to narrow 
the scope of the BoF.

SECURITY

NONE

TRANSPORT

NONE

IAB SESSIONS

NONE

OTHER SESSIONS

o Technology Deep Dives: DNS Part II
Responsible Area Director: Warren Kumari

The Technology Deep Dive session was approved for IETF 109 with no 
discussion.