+1 -- that agreement looks great
", "time": "2026-01-21T15:06:22.000Z"}, {"author": "Paul Hoffman", "text": "+1 to Carsten about getting a new draft soon.
", "time": "2026-01-21T15:07:38.000Z"}, {"author": "Rohan Mahy", "text": "there are only 5 PRs, not 16
", "time": "2026-01-21T15:07:54.000Z"}, {"author": "Rohan Mahy", "text": "but no objection to a new version
", "time": "2026-01-21T15:08:04.000Z"}, {"author": "Carsten Bormann", "text": "+1
", "time": "2026-01-21T15:09:10.000Z"}, {"author": "Mikolai G\u00fctschow", "text": "\"MUST support ordinary\" would practically restrict CBOR encoding choices retroactively, i.e. remove options available in RF8949, right?
", "time": "2026-01-21T15:17:32.000Z"}, {"author": "Carsten Bormann", "text": "Mikolai: Where that would be the outcome, we cannot do that for other peoples' protocols
", "time": "2026-01-21T15:18:13.000Z"}, {"author": "Paul Hoffman", "text": "I am successfully ignore them. We want discussions, not slides.
", "time": "2026-01-21T15:18:55.000Z"}, {"author": "Carsten Bormann", "text": "https://tzi.de/~cabo/cbor-2026-01-21-serialization.pdf
", "time": "2026-01-21T15:19:16.000Z"}, {"author": "Rohan Mahy", "text": "\n\n\"MUST support ordinary\" would practically restrict CBOR encoding choices retroactively, i.e. remove options available in RF8949, right?
\n
That would be my reading.
\nThat's why I like the SHOULD
We are using the terinology of RFC 8949 unless that is bad
", "time": "2026-01-21T15:23:14.000Z"}, {"author": "Laurence Lundblade", "text": "Yes, 8949 -- encoding/decoding
", "time": "2026-01-21T15:23:31.000Z"}, {"author": "Paul Hoffman", "text": "Fair point
", "time": "2026-01-21T15:23:36.000Z"}, {"author": "Carsten Bormann", "text": "Interoperability SHOULDs vs. RFC 6919 keywords
", "time": "2026-01-21T15:24:45.000Z"}, {"author": "Rohan Mahy", "text": "interop OR security OR ...
", "time": "2026-01-21T15:26:20.000Z"}, {"author": "Paul Hoffman", "text": "Plenty of RFCs use SHOULD/MUST for what Carsten is saying.
", "time": "2026-01-21T15:26:34.000Z"}, {"author": "Paul Hoffman", "text": "We don't need new defintions.
", "time": "2026-01-21T15:26:40.000Z"}, {"author": "Laurence Lundblade", "text": "Yes, agree that we can considered other than SHOULD in this case.
", "time": "2026-01-21T15:27:21.000Z"}, {"author": "Paul Hoffman", "text": "+1 to us saying \"good for you\" vs. \"good for interop\"
", "time": "2026-01-21T15:27:31.000Z"}, {"author": "Carsten Bormann", "text": "Giving advice is fine!
", "time": "2026-01-21T15:28:30.000Z"}, {"author": "Mikolai G\u00fctschow", "text": "can you reiterate why you treat protocol implementations separate from protocol specifications (designs)? if protocols use ordinary seralization, implementations must do so, too?
", "time": "2026-01-21T15:29:01.000Z"}, {"author": "Paul Hoffman", "text": "+1 to Laurence trying to get people to speak
", "time": "2026-01-21T15:29:02.000Z"}, {"author": "Carsten Bormann", "text": "(Doing good work with Partial implementations is an important concept)
", "time": "2026-01-21T15:29:55.000Z"}, {"author": "Carsten Bormann", "text": "CBOR definitely embraces the toolbox concept
", "time": "2026-01-21T15:31:07.000Z"}, {"author": "Carsten Bormann", "text": "Protocol implementations need protocol speciications. These do not all come from just the framework/toolbox specification.
", "time": "2026-01-21T15:32:14.000Z"}, {"author": "Ira McDonald", "text": "Once again, Meetecho has lost my one and only mic - never mind
", "time": "2026-01-21T15:32:25.000Z"}, {"author": "Rohan Mahy", "text": "feel free to make a comment here too. (and it would great to hear your comment once your mic is working again)
", "time": "2026-01-21T15:33:06.000Z"}, {"author": "Marco Tiloca", "text": "On \"Protocol implementations need protocol speciications. These do not all come from just the framework/toolbox specification.\": confirmed, thinking of EDHOC
", "time": "2026-01-21T15:33:13.000Z"}, {"author": "\u0412\u0430\u0434\u0438\u043c \u0413\u043e\u043d\u0447\u0430\u0440\u043e\u0432", "text": "zulip is not loading, meetecho does audio/video for 0.5 sec in 20 sec silence... Gov began serious internet throttling in Russia, can't speak nor hear, c ya on ML
", "time": "2026-01-21T15:33:52.000Z"}, {"author": "Carsten Bormann", "text": "Too bad, Vadim, see you on the mailing list!
", "time": "2026-01-21T15:34:29.000Z"}, {"author": "Rohan Mahy", "text": "Vadim can you comment on the contents of the slide (#6)
", "time": "2026-01-21T15:36:27.000Z"}, {"author": "Carsten Bormann", "text": "SHOULD is not useful.
", "time": "2026-01-21T15:37:00.000Z"}, {"author": "Paul Hoffman", "text": "SHOULD is useful if we say why it is not MUST
", "time": "2026-01-21T15:37:48.000Z"}, {"author": "Carsten Bormann", "text": "DLO is generally good for constrained environment.
", "time": "2026-01-21T15:37:58.000Z"}, {"author": "Carsten Bormann", "text": "It only works with MUST, though.
", "time": "2026-01-21T15:38:07.000Z"}, {"author": "\u0412\u0430\u0434\u0438\u043c \u0413\u043e\u043d\u0447\u0430\u0440\u043e\u0432", "text": "Rohan I'd change SHOULD to RECOMMENDED for last two bullets
", "time": "2026-01-21T15:39:11.000Z"}, {"author": "Carsten Bormann", "text": "Protocols MUST be explicit.
", "time": "2026-01-21T15:39:13.000Z"}, {"author": "Carsten Bormann", "text": "SHOULD == RECOMMEND in RFC 2119
", "time": "2026-01-21T15:39:30.000Z"}, {"author": "\u0412\u0430\u0434\u0438\u043c \u0413\u043e\u043d\u0447\u0430\u0440\u043e\u0432", "text": "e.g. in my merkle-canonical idea, PS is enough
", "time": "2026-01-21T15:39:56.000Z"}, {"author": "Carsten Bormann", "text": "+1
", "time": "2026-01-21T15:42:03.000Z"}, {"author": "Carsten Bormann", "text": "(MUST implement within the data model confines, with is orthogonal to serialization)
", "time": "2026-01-21T15:42:52.000Z"}, {"author": "Carsten Bormann", "text": "which
", "time": "2026-01-21T15:42:58.000Z"}, {"author": "Rohan Mahy", "text": "+1 to what Joe just said
", "time": "2026-01-21T15:43:21.000Z"}, {"author": "Joe Hildebrand", "text": "No. MUST implement everything, because otherwise you can't check the data model correctness.
", "time": "2026-01-21T15:43:30.000Z"}, {"author": "Carsten Bormann", "text": "OK, interesting: erroring out on a floating point value if the data model doesn't have them is a valid implementation strategy
", "time": "2026-01-21T15:44:10.000Z"}, {"author": "Mikolai G\u00fctschow", "text": "+1 to say \"SHOULD say something\", instead of imposing ordinary
", "time": "2026-01-21T15:44:18.000Z"}, {"author": "Carsten Bormann", "text": "Ordinary (if that is DLO+PS) is not a good MUST enforce in the decoder.
", "time": "2026-01-21T15:45:34.000Z"}, {"author": "\u0412\u0430\u0434\u0438\u043c \u0413\u043e\u043d\u0447\u0430\u0440\u043e\u0432", "text": "+1, not MUST
", "time": "2026-01-21T15:45:56.000Z"}, {"author": "Mikolai G\u00fctschow", "text": "re Joe, Carsten above: MUST not implement fully but, \"everything that the data model allows\"
", "time": "2026-01-21T15:47:12.000Z"}, {"author": "Ira McDonald", "text": "Sorry no mic function here today
\nLL's example of COSE (no one implements it all, it's not a protocol like TLS) is a relevant example - since CBOR is never a protcol itself.
Yes, COSE is different than TLS. TLS and FIDO are similar in this way.
", "time": "2026-01-21T15:49:47.000Z"}, {"author": "Carsten Bormann", "text": "TLS is a protocol in the sense of sending back and forth messages; CBOR is a data representation format -- that is bound to be different
", "time": "2026-01-21T15:50:54.000Z"}]