Re: [quicwg/base-drafts] Padding outside QUIC packet (#3333)

[Kazuho Oku <[e-mail]>]

This behavior should continue to be allowed. We allow endpoints to send a QUIC packet (as part of a coalesced packet) knowing that the peer would de unable to decrypt it (see [section 14.3.1](https://quicwg.org/base-drafts/draft-ietf-quic-transport.html#section-14.3.1)). The issue being discussed here is a variant of that.

That said, I am not sure if we want to suggest padding outside of a QUIC packet, because such a design might have wired implications to the CC logic. IIUC, current congestion logic is based on the following principles:
* an ack-eliciting packet consumes bytes-in-flight
* an ack-only packet does not consume bytes-in-flight
* we allow ack-eliciting packets and ack-only packets to be coalesced, because that does not change the load on the network

I am not sure how we can adjust these principles so that padding outside QUIC packets can be taken into consideration (when necessary), or if making such adjustment is worth the additional complexity.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/3333#issuecomment-573952468