Ballot for conflict-review-aranda-dispatch-q4s
Yes
No Objection
Note: This ballot was opened for revision 00 and is now closed.
Ballot question: "Is this the correct conflict review response?"
Balloting Yes, since there's no conflict with IETF work. While reviewing, I did notice some areas where the document could perhaps use more polish: It discusses handling the "user's location" or updates thereto, as well as carrying an explicit "public-address" attribute, but does not discuss the usability of IP addresses as PII and the corresponding privacy considerations. UTF-8 is used for human-readable strings but I didn't see mention of normalization or other internationalization considerations. There's a 505 "Version not supported" error code but not much of a clear mechanism for actually performing version negotiation. The default max content length of 1000 bytes doesn't match up with any natural transport limits I'm aware of, and in fact is larger than an IPv4 network is guaranteed to be able to carry. Similarly, when we read that "all BWIDTH requests sent MUST be 1 kilobyte in length (UDP payload length by default" it's unclear how that would work on a network with MTU smaller than 1kb, and what the "UDP payload length by default" refers to. Section 9.6.1 talks about cases when "TCP packets are lost" but TCP is a stream protocol, not a packet-based one, and more relevantly, provides reliable transport, so loss of carried TCP data is detectable only as connection hang/timeout.
No objection, however, I'd be curious how this document got to the ISE. Was this ever presented at dispatch (as the name indicates) or somewhere? Does anybody now?