Re: [quicwg/base-drafts] Pad path validation in both directions (#4241)

[Jana Iyengar <[e-mail]>]

@janaiyengar commented on this pull request.



>  frame unless constrained by congestion control.
 
+A PATH_RESPONSE frame SHOULD be sent on the network path where the
+PATH_CHALLENGE was received.  This ensures the path is functional in both
+directions.  This requirement MUST NOT be enforced by the endpoint that
+initiates path validation as that would enable an attack on migration; see
+{{off-path-forward}}.
+
+An endpoint SHOULD expand datagrams that contain a PATH_RESPONSE frame to at

A SHOULD suggests there are potentially good reasons where an implementation might deviate from the recommendation. If there aren't any, then it's a MUST. It doesn't have to be enforceable, since not doing it means the protocol might not work as intended. That is the case here -- not following the MUST means that the protocol might not work.

In terms of the value in deviating from this recommendation, I don't see this degree of freedom being useful for multipath, though it's likely I'm not seeing what you're thinking. That said, I don't think we should design multipath here.

-- 
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/pull/4241#discussion_r508195566