[Cbor] Re: CBOR Tooling breadth of use

[Steve Lasker <[e-mail]>]

Thanks, Carsten, A.J.

Sounds like the CBOR WG would like to see broad adoption, and is willing to assist, we just need folks to ask. Thank you!

There are many tools to enable, VSCode is one of many. Orie has written the beginnings of a VSCode extension (https://github.com/transmute-industries/vscode-scitt-preview) , but needs more help (or time from someone that can take it to completion).



A.J. – yes, round tripping a serialization format is “easy”-ish. There are many aspects that go beyond. Auto-complete, validation, are the generic ones. Pulling in public and private CBOR registries are another specific to CBOR.



I'll throw my hat in helping scope, build, document, evangelize CBOR and COSE tooling

Without spamming this WG further, If anyone on this mailing group would like to contribute, please let me know and I'd be happy to help facilitate.

From: A.J. Stein <[e-mail]>
Sent: Thursday, July 25, 2024 5:30 PM
To: Steve Lasker <[e-mail]>
Cc: [e-mail]
Subject: Re: [Cbor] CBOR Tooling breadth of use

On Thu, Jul 25, 2024 at 6:53 PM Steve Lasker <[e-mail]<mailto:[e-mail]>> wrote:
Working with developers outside of IETF, adoption of CBOR has met resistance, …due to a lack of tooling. There have been various libraries starting to surface, for using CBOR in code, however the lack of visualization, editing, debugging is a limitation. It’s just too easy to open a json, yaml, (dare I say xml) files. To pass parameters, et.

It is easy to create software to deserialize and serialize "human-friendly" markup languages and JSON (the others precisely have "ML" in their initialism; is JSON one of them it depends on how you use it) that can deny service to a system and be the primary or significant cause to compromise. I do not want to make the email all about that, but I can compile a list of issues around that.

How much energy and acknowledgement is there around CBOR and COSE tooling? Does the CBOR community feel it’s not intended as a general-purpose data format, and it’s good enough? Or, is there a desire to make easier for more to use?

You are part of this community by virtue of emailing. So we have energy and acknowledgement with you, I, and Carsten just responded. What is good enough? What do you desire? What can you I or commit to move the community forward?