Re: [Cbor] Paul Wouters' No Objection on draft-ietf-cbor-time-tag-11: (with COMMENT)

[Carsten Bormann <[e-mail]>]

Hi Paul,

thank you for these comments.

I started a PR [1], but expect more IESG reviews to come in so we’ll collect those before submitting a -12.

[1]: https://github.com/cbor-wg/time-tag/pull/22

Grüße, Carsten



> On 2023-10-23, at 20:35, Paul Wouters via Datatracker <[e-mail]> wrote:
> 
> The document, along with other ballot positions, can be found here:
> https://datatracker.ietf.org/doc/draft-ietf-cbor-time-tag/
> 
> 
> 
> ----------------------------------------------------------------------
> COMMENT:
> ----------------------------------------------------------------------
> 
> NITS:
> 
> TAI is never expanded - expand on first use? At least for me this was a new
> term, unlike UTC.

(Which also is never expanded — in both cases because these are actually not acronyms.)
Time-nuts (*) know what TAI and UTC is :-)
Given that we already reference draft-ietf-sedate-datetime-extended, which has a definition for UTC and at least an expansion for TAI, we could simply point there.

I added:

>> Several time-related terms such as UTC and TAI are discussed in
>> {{IXDTF}}, which may be a useful companion document beyond its direct
>> use in Sections {{<tzh}} and {{<suff}}.

Now
1b83711

> But perhaps it is obvious for implementers in this space.

I would expect UTC to be entirely obvious at this stage (even if people maybe don’t know the exact definition), while TAI is probably more specialized.

>        The present document
> 
> Use "This document"

I’m hearing this kind of change request often; I don’t know why.  The text here talks about several documents, and “this document” then is an ambiguous reference.  But the RFC editor will make some good proposals.

>        therefore is rendered by the surrogate notation seen here in the
>        plain-text rendition.
> 
> "here" is a bit weird when I'm reading the HTML version. Can't you just write
> 2^xx as an example ?

We honestly don’t know how 2<sup>xx</sup> is rendered in plaintext.

IIRC, the details have changed twice over the last five years (see Section 1.2 of RFC 8949 for one of the intermediate states, and Section 1.1 of RFC 9100 for another).
The weird way this is said in the text here has been conditioned by this experience.

(There are some fundamental problems with rendering processes that produce multiple formats, in particular when the text needs to be self-referential…)

>        The security considerations of RFC 8949 apply
> 
> A link is missing here.

Now
715ddcf



(*) Time-nut: Somebody who has multiple atomic clocks and GNSS-disciplined oscillators and knows what an Allan deviation is.