[{"author": "Michael Richardson", "text": "

Hello

", "time": "2024-03-18T07:32:48Z"}, {"author": "Brian Carpenter", "text": "

Esko's hand is up too

", "time": "2024-03-18T07:40:14Z"}, {"author": "Michael Richardson", "text": "

Sorry, Yes. my USB comments were about BRSKI-PRM. HOWEVER, to make things work over USB key, it needs to use FULLCMC.

", "time": "2024-03-18T07:52:12Z"}, {"author": "Toerless Eckert", "text": "

andmoved the comments in the notes down to PRM section ;-)

", "time": "2024-03-18T07:52:45Z"}, {"author": "Michael Richardson", "text": "

RFC9417 and RFC9418. Service Assurance for Intent-Based Networking Architecture

", "time": "2024-03-18T08:11:13Z"}, {"author": "Steffen Fries", "text": "

Regarding the USB approach. I don't see the need for FULLCMC in hat context, as PRM also uses the additional wrapping signature

", "time": "2024-03-18T08:12:20Z"}, {"author": "Steffen Fries", "text": "

may also discuss this in Design Team next week

", "time": "2024-03-18T08:13:41Z"}, {"author": "Michael Richardson", "text": "

Steffen Fries said:

\n
\n

Regarding the USB approach. I don't see the need for FULLCMC in hat context, as PRM also uses the additional wrapping signature

\n
\n

Hmm. Like I said: I want to be sure I know what I'm talking about, and I'll 1:1 with you to figure this out.

", "time": "2024-03-18T08:21:33Z"}, {"author": "Michael Richardson", "text": "

What fixed width fields does GRASP have? If it's all CBOR, one can use 1-byte integers (-23 to +24)?

", "time": "2024-03-18T08:24:29Z"}, {"author": "Michael Richardson", "text": "

I agree that the mix UDP/TCP is a bit weird for many.

", "time": "2024-03-18T08:24:58Z"}, {"author": "Michael Richardson", "text": "

My feedback: I need more details.

", "time": "2024-03-18T08:25:41Z"}, {"author": "Michael Richardson", "text": "

(about lightweight GRASP)

", "time": "2024-03-18T08:26:04Z"}, {"author": "Michael Richardson", "text": "

[I think the idea is to run over QUIC[UDP] rather than move it all to TCP. So it's not UDP->TCP, it's getting rid of the TCP]

", "time": "2024-03-18T08:28:05Z"}, {"author": "Thomas Werner", "text": "

Michael Richardson said:

\n
\n

Steffen Fries said:

\n
\n

Regarding the USB approach. I don't see the need for FULLCMC in hat context, as PRM also uses the additional wrapping signature

\n
\n

Hmm. Like I said: I want to be sure I know what I'm talking about, and I'll 1:1 with you to figure this out.

\n
\n

Great Michael, lets have a closer look at the use case you mentioned ... :-)

", "time": "2024-03-18T08:32:23Z"}, {"author": "Michael Richardson", "text": "

We could think about replacing TCP with CoAP w/OSCORE.

", "time": "2024-03-18T08:33:10Z"}, {"author": "Michael Richardson", "text": "

DTLS can not do multicast, but CoAP can.

", "time": "2024-03-18T08:33:40Z"}, {"author": "Michael Richardson", "text": "

and group OSCORE too.

", "time": "2024-03-18T08:34:01Z"}, {"author": "Carsten Bormann", "text": "

+1

", "time": "2024-03-18T08:34:15Z"}]