Nice new milestone for #embedded #Rust development: defmt 1.0 has been released. The crate enables logging for microcontrollers without text or number formatting on the device – instead, the device sends numbers of interned strings to the host, which composes log messages.
1.0 is not only a stable release, but also enables logging #CBOR data into diagnostic notation on the host.

The #IETF122 hackathon starts in a few hours. I will be joining remotely, work on #EDHOC on @ariel (let's see if its out-of-the-box support also interoperates out-of-the-box), and play with #embeddfriendly URIs expressed in #CBOR.

RFC 9741: CDDL Additional Control Operators for the Conversion and Processing of Text

Le langage #CDDL, qui permet de créer un schéma formel pour des formats comme #CBOR, peut s'étendre via l'ajout d'« opérateurs de contrôle ». Ce #RFC en spécifie quelques uns, notamment pour agir sur du texte ou convertir d'une forme dans un autre.

https://www.bortzmeyer.org/9741.html

My recent work on #CBOR diagnostic tools has culminated in the release v0.0.7 of the cbor-edn #RustLang crate, and the cbor-diag v1.1.0 #Python package (which is powered by the former, #maturin and #PyO3).
It is now easy to receive CBOR data structure with internal tags, and have them display like this:
{1: IP'2001:db8::1', 3: DT'2025-01-14T13:27:49Z'}
https://cbor-diag.readthedocs.io/en/latest/
https://docs.rs/cbor-edn/latest/cbor_edn/

RFC 9682: Updates to the CDDL grammar of RFC 8610

Voici une légère mise à jour de la grammaire du langage de description de schéma #CDDL (Concise Data Definition Language), originellement normalisé dans le RFC 8610. Pas de gros changement.

https://www.bortzmeyer.org/9682.html

Starting into a busy #IETF121 day: talking about #CoAP in the morning CoRE meeting, IoT lunch, #CBOR in afternoon, and board game night to unwind.
And that's only the Tuesday :-)

This week, I am at the @ietf Hackathon and #IETF121 in Dublin. In #CoRE and #CBOR WGs, I will present #DNS over #CoAP and related work to make DNS more concise and resource-friendly. Check our drafts at https://datatracker.ietf.org/person/martine.lenders@tu-dresden.de

On my way to the #IETF121 hackathon -- I'll be working on packed #CBOR and #CoAP security, but also curious to chat with folks from other projects.

The #RustLang #CoAP example of @RIOT_OS has been updated. Previously you could list processes, access sensors or read the file system. Now, you can blink LEDs, read network settings, send #IPv6 pings, and even see who pinged you.
The data formats will still need some work, right now a lot is just crude #CBOR, but it works.
https://github.com/RIOT-OS/RIOT/tree/master/examples/rust-gcoap

RFC 9581: CBOR Tags for Time, Duration, and Period

Ce #RFC ajoute au format de données binaire #CBOR la possibilité de stocker des données temporelles plus détaillées, incluant par exemple l'échelle utilisée (UTC ou TAI), ou bien ayant une précision supérieure.

https://www.bortzmeyer.org/9581.html

yeah, my cbor writer can handle `std::ranges::views`

It's been on my mind I wanted to get to understand a few things…

1. #COSE : #CBOR Object Signing and Encryption
2. #X25519KeyExchange

There are a few applications I have in mind for this and I could find no good examples. So, I sat down and nutted out this:

https://gist.github.com/sjlongland/fc2e06cb47543e977440b79d6cf71185

This is a simple demonstration of X25519 and COSE in #Python. The idea being that this allows two peers to mutually authenticate, then establish a shared secret.

I'm still not 100% certain of some X25519 key exchange details, specifically why pycryptography's example does it twice?

https://cryptography.io/en/latest/hazmat/primitives/asymmetric/x25519/#exchange-algorithm

Note: they generate `derived_key`, but never use it. It'd make sense if it were part of the second call to `HKDF` or something, but the result is simply thrown away. Anyone know why that is?

Anyway, maybe the #cryptography gurus might have some ideas, and whether my little toy here actually would work for decent secure comms, or if there's a gaping hole I should look out for.

Getting out of a sidetrack activity: cbor-edn 0.0.3 published. It is both a command line tool and a #RustLang library, and interacts with #CBOR diagnostic notation from plain conversion up to processing of the upcoming application literals:
$ cargo install cbor-edn
$ echo '{4: 1234567890}' | cbor-edn diag2diag --annotate ccs
{/ exp /4: dt'2009-02-13T23:31:30+00:00'},
https://codeberg.org/chrysn/cbor-edn

Are you using #CBOR in data formats you describe, in code, debugging or for examples? During the last #IETF meeting, @cabo gave a 30 minute tutorial on existing and upcoming technologies to make that easier. Watch the recording at <https://youtu.be/Cr6PeUUm7wY?feature=shared&t=237>; slides with further links <https://datatracker.ietf.org/meeting/119/materials/slides-119-cbor-edn-cddl-yang-packed-00>.

Gathering #CBOR #RustLang libraries for embedded use, early 2024 edition:
* minicbor looks fine; its derive macros take some getting used to
* tps_minicbor looks powerful, but insists on memmoves to allow serializing iterators w/o known lengths
* ciborium is powerful but deeply alloc based; its ciborium-ll subset that is no-alloc is meek.
* cbor4ii is an active successor to serde_cbor
The latter two suffer serde's trouble with CBOR formats not resembling its mental model.

Connaissez-vous la RFC 8949 ? Non ? Jusqu'à ce matin, à ma grande honte, moi non plus... Pourtant le sujet est d'importance : une alternative binaire, compacte, performante, normée et pérenne. Le #CBOR: Concise Binary Object Representation.
https://cbor.io/

Le seul tuto du site renvoie à un article fr de présentation de @bortzmeyer : https://www.bortzmeyer.org/7049.html

#BSON, #protobuf, #MessagePack : chacun a ses avantages (et inconvénients) face à #JSON.
Le CBOR est une couleur de cette palette.