Re: [Gen-art] Genart last call review of draft-ietf-cbor-cddl-control-05

[Theresa Enghardt <[e-mail]>]

Hi Carsten,

Thanks for the reply and changes, they look good to me.

On 9/17/21 8:28 AM, Carsten Bormann wrote:
>> Section 7 (Security considerations):
>> Can there be any additional security concerns if CDDL specifications can
>> contain ABNF or "arbitrary" features? While this document obviously can't go
>> into specifics, maybe it's worth calling out that one needs to pay specific
>> attention if these control operators are used.
> Isn’t that true for any extension to any language?
> I’d prefer to highlight specific security considerations when I’m aware of them.

Probably true, but I guess I prefer to state the obvious, especially 
when it comes to security, to be safer rather than sorry. Also I like it 
when a document that "builds" on another document still shows that the 
authors considered the implications of what they added.

That said, I won't argue with your reasoning :)


>> Nits/editorial comments:
>>
>> Section 2.2
>> "concatenating the target text string ""foo"""
>> Is foo intended to be in two double quotes, or should there only be one pair of
>> quotes?
> This is an artifact of the way xml2rfc handles <tt> elements.
> There currently is no way to mark the document up in a way such that both plain text and html versions look good, and I opted to prefer the HTML to look good (and correct).
> (RFC 8949 has a long explanation in several paragraphs of Section 1.2 how the plain text sometimes will look weird; I was hoping I wouldn’t have to repeat this here.)
> There is a proposed change to xml2rfc that will change the specifics, so I would like to wait for how that plays out before making changes.
>
>
>> Section 3
>> "by defining a ".abnf" control operator"
>> Should this say 'an ".abnf" control operator' instead?
> Actually, we say out loud “dot a b n f control operator”, so “a” corresponds to the speech pattern.

Thanks, I learned something in both of these cases.

Best,
Theresa