Minutes IETF123: icnrg: Fri 07:30
minutes-123-icnrg-202507250730-01

ICNRG Meeting
IETF Madrid - July 25 2025

Introduction

Dirk Kutscher will not continue as INCRG chair now that he leads the
IRTF. Thanks Dirk for everything!

Clock synchronization for defined-trust Limited Domains - Kathleen
Nichols, Pollere

Working in a trust domain - TD (picture on the slides)
4 interations is working well for the calibration between neighbors
Testing results for 8 fully connected (via broadcast) members.
ICN changes the perspective

Question: (Dave Oran) How to backmap the virtual clock onto the system
clock?
Answer: Virtual clock to stamp the publication - timer is via the
system clock
Questions: (David Venhoek) 1. You seem to use timing that is Larger
than the RTT how does this influence the stabilization of the system 2.
At what point the clock is stabilized/synchronized
Answers: 1. You should have set your tolerance to be larger - looking
forward to have a larger testbed offered by other people 2. Slides give
numbers - 12 s is a number that needs to be verified

ICN for Cybernetic Avatar and Metaverse - Yusaku Hayamizu & Hitoshi
Asaeda, National Institute of Information and Communications Technology
(NICT)

Developed use cases in remote work etc.
Activities in ITU about ICN and Metaverse

Comment (Dave Oran): I put a suggestion in the chat - Suggest though
you change the abbreviation for the robot from "CA" to something else -
everybody in IETF world uses that to mean "certificate authority"

AT Protocol - Daniel Holmgren - BlueSky

AT presented to be brought in the IETF (the data repository) - atproto
Distributed protocol
Different from W3C ActivityPub
Based on CBOR (derived from CoAP)
Decentralized and distributed data repository

Question: (Dave Oran)
90% of people chose their name under the same domain as to not have to
deal with DNS - do you see a better way to get social network
identities?**
Answer: Community run handles is a way to get more diversity. DNS is
getting more used even by non-technical folks.
Question (Dave Oran) How are those mapping DNS are mapped? Certs?
Answer: We do not use certs.
Question (Hitoshi Asaeda) Do you have any thoughts on implementing
this for ICN? You could profit from ICN technically.
Answer: It is not at the networking layer.
Comment (Hitoshi Asaeda): Recently, many ICN studies have run ICN on
overlay networks. Cefore and NDN can run on top of IP. The network layer
does not necessarily need to be changed. ICN is advantageous even on IP
networks.

Question: (Mohammed Diallo) Martin Kleppman is the grand sheik of
distribution system. Limit of the size of Bluesky ID - can it be
changed? Latin vs Arabic character.
Answer: Put a bug report.
Comment: (Dirk Kutscher) The reason ICNRG invited Daniel is because of
some conceptual similarities. In ICN there had been some similar
concepts but this works globally.

1. Chunking Update - Hitoshi Asaeda

New version posted yesterday - commenting via the mailing list.

1. Closing and next steps

Need people to review the current drafts to push them expedititely
towards RFC: FLIC (been there long time) and Reflexive Forwarding.
Also: CCXChunking ready for last call?
Meeting in Montreal: arrive on the 30th for Halloween!