Re: [Cbor] CDDL 2 -- non-literal tags support

Carsten Bormann <cabo@tzi.org> Tue, 07 March 2023 14:08 UTC

Content-Type: text/plain; charset="utf-8"
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.120.23.2.7\))
From: Carsten Bormann <cabo@tzi.org>
In-Reply-To: <DB9PR08MB65249ABDEB8DFFC5E3F5E1489CB79@DB9PR08MB6524.eurprd08.prod.outlook.com>
Date: Tue, 07 Mar 2023 15:08:47 +0100
Cc: "cbor@ietf.org" <cbor@ietf.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <459BF171-857D-4A4F-BE3A-7C1CA31925B5@tzi.org>
References: <DB9PR08MB65249ABDEB8DFFC5E3F5E1489CB79@DB9PR08MB6524.eurprd08.prod.outlook.com>
To: Thomas Fossati <thomas.fossati@arm.com>
Archived-At: <https://mailarchive.ietf.org/arch/msg/cbor/56wOtzbJ3kalnLrUAy17WuEyGHo>
Subject: Re: [Cbor] CDDL 2 -- non-literal tags support
Precedence: list

On 2023-03-07, at 14:38, Thomas Fossati <thomas.fossati@arm.com> wrote:
>
> If I were there though, I'd be rooting for non-literal tags [1] to be finalised asap.

Right. At the moment, I see the following steps, probably each with a document on its own:

* Section 2 of draft-bormann-cbor-cddl-2-draft, except for 2.3

These are grammar fixes that are conceptually well-understood and at least partially implemented.
This would be one component that I’d like to finish very soon as an update to RFC 8610, maybe informally calling it CDDL 1.1 (because these are just a few little fixes).

* draft-bormann-cbor-cddl-more-control

These are additional control operators that simply exercise the existing extension point, like RFC 9165 did. Most of the ones in this set are fully implemented. We are in the process of getting some more feedback from implementers using those. I hope to have an update of the spec ready before the IETF 116 I-D deadline. This also could go ahead quickly, forming a triad with the above and RFC9165 and forming a new baseline for CDDL 1 support.

* Section 4 of draft-bormann-cbor-cddl-2-draft (“modules”)

A first draft implementation is out there. This could be the first step of CDDL 2.0. A possible later extension would be Appendix A.2, but that can be a separate document (on a longer timeline as it involves additional entities such as IANA) — it is not required to finish the current Section 4.

* Section 3 of draft-bormann-cbor-cddl-2-draft

Lots of music, lots of fun. I think we will be playing a lot with this (starting at the IETF 116 Hackathon) before we fully write it up. Could be something like a CDDL 2.5. Until then, .feature already helps some.

Not on the main line of development, but still of interest:

* Section 2.3/A.1 of draft-bormann-cbor-cddl-2-draft, in conjunction with draft-bormann-cbor-edn-literals
* what else is in draft-bormann-cbor-cddl-freezer (e.g., I think we could document the JSON format for tool interchange soon, after CDDL 1.1)
* draft-bormann-cbor-rfc-cddl-models
* draft-bormann-cbor-draft-numbers
* draft-bormann-cbor-cddl-csv

Of course, this list interacts with documents outside the CDDL stream that also do carry CDDL (e.g., time-tag and CMW).

Grüße, Carsten

[Cbor] CDDL 2 -- non-literal tags support Thomas Fossati
Re: [Cbor] CDDL 2 -- non-literal tags support Carsten Bormann