Re: [Cbor] [Cacao] Public Comment for CACAO Draft

[Carsten Bormann <[e-mail]>]

Hi Bret,

On 2020-12-15, at 21:38, Bret Jordan <bret.jordan=[e-mail]> wrote:
> 
> Or can we get a working group for this kind of work.

I would consider a document that defines vocabulary like this to be a good component for CDDL, which the CBOR WG already is chartered to evolve.  A bit like a modern version of the good parts of part 2 of WSD [1], which is probably still the best document of that kind that we have.

(Clearly, I’m not suggesting that CACAO wait for this to spring up, I just see the CACAO document as another item to look at that motivates finally doing this.)

Grüße, Carsten

[1]: https://www.w3.org/TR/xmlschema-2/

(Oh, and the difference between a 54-bit two's-complement signed integer and an I-JSON integer is that the former includes an INT54_MIN of -2**53, while the latter doesn’t (*), which you correctly copied from I-JSON — int54 may be vernacular for the latter, but this is off-by-one.  This little factoid is pretty much inconsequential (**) for CACAO, but a document like the above would get this detail right.)

(*) because it actually does include both +0 and -0.
(**) INT54_MIN *can* be represented in the typical JSON implementation number systems that I-JSON was accommodating, as can be 2**53 — it’s just no longer uniquely distinguishable from INT54_MIN-1 or 2**53+1.  So this little mistake is even more innocuous.