On 09/11/2022 09:44, Jonathan Morton wrote:
On 9 Nov, 2022, at 11:10 am, Bob Briscoe <i...@bobbriscoe.net> wrote:
2/ Another question: Why 'and Signalling' in the title?
Other than the title and response to congestion signals, the charter doesn't
say anything about signalling itself.
If, say, there were more work on tunnelling ECN or something, would tsvwg take
that on, or congress?
"Congestion signalling" refers not only to the mechanisms that allow congestion
to be detected by transports, but the algorithms (from drop-tail overflow upwards) which
generate those congestion signals to control transports' demand on network resources.
The former aspect, referring to the wire protocols, may well be properly handled by TSVWG
and/or individual protocol WGs (eg. TCPM, QUIC, etc). The latter aspect, however, is more
general in terms of protocol and more specific to congestion control itself. I think "…
& Signalling" would refer very naturally to text mentioning AQM in the body of the
charter.
- Jonathan Morton
Clearly, if an AQM is designed only for a constrained environment, and
never intended for the general Internet, placing this with transport CC
might work - of course, *IF* the IETF decides that protocols for
constrained environment is something where standards are important.
After thinking further on this though, I do think that new AQM deisgns
ought to take into consideration other forwarding at the IPv4/IPv6
packet level, this is something that needs careful coordination with
other aspects of forwarding (as in RFC 7567). Therefore, I do not
something where we can simply encourage an experiment with a range of
behaviours, and I do not think it should be put in a separate group.
Gorry
Gorry
