I suggest we hold this for update.  This is not necessary (a MAY and MUST are 
not in conflict) and it complicates the section to mention NEW_TOKEN this way, 
but it's not outright wrong.

On Fri, Jun 12, 2026, at 04:06, [email protected] wrote:
> The following errata report has been submitted for RFC9000,
> "QUIC: A UDP-Based Multiplexed and Secure Transport"
>
> --------------------------------------
> You may review the report below and at:
> https://errata.rfc-editor.org/eid9003/
>
> --------------------------------------
> Type: Technical
> Reported by: zhangph <[email protected]>
>
> Section 12.5 says:
>
> Original Text
> -------------
> Note that it is not possible to send the following frames in 0-RTT
> packets for various reasons: ACK, CRYPTO, HANDSHAKE_DONE, NEW_TOKEN,
> PATH_RESPONSE, and RETIRE_CONNECTION_ID.  A server MAY treat receipt
> of these frames in 0-RTT packets as a connection error of type
> PROTOCOL_VIOLATION.
>
> Corrected Text
> --------------
> Note that it is not possible to send the following frames in 0-RTT
> packets for various reasons: ACK, CRYPTO, HANDSHAKE_DONE, NEW_TOKEN,
> PATH_RESPONSE, and RETIRE_CONNECTION_ID.  A server MAY treat receipt
> of these frames in 0-RTT packets as a connection error of type
> PROTOCOL_VIOLATION.  Note that receipt of a NEW_TOKEN frame is always
> a connection error of type PROTOCOL_VIOLATION, as specified in
> Section 19.7, and is not subject to the MAY above.
>
> Notes
> -----
> Section 19.7 states unconditionally that "a server MUST treat receipt 
> of a NEW_TOKEN frame as a connection error of type PROTOCOL_VIOLATION", 
> because NEW_TOKEN is a server-to-client frame. Section 12.5 lists 
> NEW_TOKEN among the frames for which a server MAY treat receipt in 
> 0-RTT packets as a connection error. For the specific case of a server 
> receiving a NEW_TOKEN frame in a 0-RTT packet, the permissive MAY in 
> Section 12.5 reads as if it could override the unconditional MUST in 
> Section 19.7. While Section 19.7 (being specific to NEW_TOKEN) would 
> normally govern, the spec does not explicitly state this precedence, 
> leaving an apparent normative-strength overlap. This is submitted as 
> Editorial: the proposed note in Section 12.5 clarifies that Section 
> 19.7's MUST takes precedence for NEW_TOKEN in all packet number spaces, 
> removing the ambiguity. (No change to intended behavior is requested, 
> only clarification of the relationship between the two sections.)
>
> Instructions:
> -------------
> This erratum is currently posted as "Reported". Please
> use "Reply All" to discuss whether it should be verified or
> rejected. When a decision is reached, the verifying party  
> will log in to change the status and edit the report, if necessary.
>
> --------------------------------------
> RFC9000 (draft-ietf-quic-transport)
> --------------------------------------
> Title               : QUIC: A UDP-Based Multiplexed and Secure Transport
> Publication Date    : May 2021
> Author(s)           : J. Iyengar, Ed., M. Thomson, Ed.
> Category            : Proposed Standard
> Source              : quic (wit)
> Stream              : IETF
> Verifying Party     : IESG

Reply via email to