Hi all, We are running a project at the IETF 126 Hackathon (18-19 July, Vienna) on a minimal multicast extension for QUIC.
Motivation: draft-navarre-quic-flexicast defines Flexicast QUIC, which supports multicast delivery, builds on Multipath QUIC, and requires changes across several parts of the stack. draft-jholland-quic-multicast takes a different approach, focusing on the security aspects of the extension. For the hackathon, we want to explore the lower bound of the design space: what is the smallest set of changes to QUIC that still yields working multicast delivery, and what do we give up to get there? We expect the outcome to be multiple proof-of-concept extensions in multiple existing implementations, plus a clearer picture of which parts of Flexicast QUIC are essential and which are conveniences. Project repository: https://github.com/louisna/minimal-multicast-quic-ietf-126 If you are interested, please reach out or open an issue on the repo. Relevant drafts: - draft-navarre-quic-flexicast: https://datatracker.ietf.org/doc/draft-navarre-quic-flexicast/ - draft-jholland-quic-multicast: https://datatracker.ietf.org/doc/draft-jholland-quic-multicast/ Scientific paper: Taking the Best of Multicast and Unicast with Flexicast QUIC: https://dl.acm.org/doi/pdf/10.1145/3750832.3750834 Best regards, Louis Navarre, UCLouvain
