Investigate QUIC and the unreliable datagram extension for peer to peer

[zmanian]

The IETF QUIC (a UDP based protocol which combines some characteristics of connections and TLS) is also potentially having datagram semantics.

It also includes a datagram semantics over the same connection.

This will eliminate RTT overhead on sending votes to peers.
https://datatracker.ietf.org/doc/draft-pauly-quic-datagram/?include_text=1

[thanethomson]

Since there's been talk of moving to libp2p, would the best way to go about doing this be to contribute an implementation of the draft spec to go-libp2p-quic-transport? Or would we want to create our own implementation from scratch?

[liamsi]

Speaking of libp2p: I heard rumours about adding something more noise like to libp2p (I could not find any spec etc yet though).
Some background https://forum.web3.foundation/t/transport-layer-authentication-libp2ps-secio/69

[tac0turtle]

Should issues like these be moved to the spec repo for further investigation?

[cmwaters]

QUIC is being considered as part of the LibP2P migration so I think it's fine to close this issue. More information can be found in this ADR