Re: [Ntp] [EXT] Re: Fwd: Antwort: Re: WGLC - draft-ietf-ntp-ntpv5-requirements

Thank you for your feedback, responses in line. Changes have been applied to the editors copy in Github, a new version of the document will be published later when more dust settles.

> On 20 Dec 2023, at 17:27, Ira McDonald <blueroofmusic@gmail.com> wrote:
> 
> Hi,
> 
> I have reviewed this document and consider it almost ready for IETF Last Call, with the following comments:
> 
> 1) Title: As an Informational document, the title should be changed as Ulrich has suggested to perhaps
>     "Design Considerations for NTPv5" or "Design Considerations for NTPv4 Successors" (w/ a nod to Harlan)
>     - keeping "Requirements" in the title is not acceptable in an Informational document

I disagree and see no benefit at this stage of renaming the document; the title of the document is unambiguous, and removing reference to NTPv5 given that there is already a working group document defining the NTPv5 protocol does not appear beneficial. This is not the first nor last informational document that describes requirements, and I'm unaware of any rules or guidelines that prohibit it explicitly, unless you are aware of one that I have overlooked?

> 2) Keywords: As an Informational document, the keywords "MUST" and "SHALL" should never be used
>     and instead should be globally replaced with "SHOULD"

Some of those MUSTs (e.g. section 4.1 "Client and server protocol modes MUST be supported...") are the result of extensive working group consensus calls and I am reluctant to modify them at this point. If there are specific suggestions you have I'd be open to adjusting the document.

> 3) Keywords: Multiple instances of lowercase "should" (e.g., section 3.1, paragraph 2) appear to be intended
>     to be "SHOULD" (a recommendation)
>     - all lowercase uses of "should" should be replaced with either "SHOULD" or "can/could/" (consistent with
>     ISO editorial rules)

I'm glad to receive any specific feedback and suggestions that you have.

> 4) Keywords: Multiple uppercase keywords should *never* be used in a single sentence (e.g., section 4.1,
>     paragraph 2), because this style adds inherent ambiguity to the recommendations.

I've reviewed this example and several others which may be ambiguous, however as per above some normative text is the result of agreed text through working group consensus so I'm reluctant to do any heavy structural changes. That said, if any that remain you believe are ambiguous please tell me which specific examples are and they can be addressed. I see little benefit in breaking up sentences that are clear enough just to stick to this.

> 5) Leap Seconds: I suggest that a discussion of Leap Seconds and Smearing should be included in this
>     document - the current text seems reasonable to me.
> 
> 6) Backwards Compatibility w/ NTPv4: I strongly agree that NTPv4 compatibility is not important.
>     - core Internet protocols (such as NTP) evolve over time and explicit version fields are the right path.
> 
> Cheers,
> - Ira
> 
> 
> On Tue, Dec 19, 2023 at 3:46 AM Windl, Ulrich <u.windl@ukr.de> wrote:
> Hi!
>  Maybe once the NTPv5 spec is final we can rename the requirements to an informative RFC titled “considerations defining NTPv5” or similar. So maybe some interesting considerations that did not make it in the official spec would be preserved for readers with historical interest. I’m not deeply in the RFC making process, however.
>  Regards,
> Ulrich
>   From: ntp <ntp-bounces@ietf.org> On Behalf Of Salz, Rich
> Sent: Thursday, December 14, 2023 6:29 PM
> To: Daniel Franke <dfoxfranke@gmail.com>; NTP WG <ntp@ietf.org>
> Subject: [EXT] Re: [Ntp] Fwd: Antwort: Re: WGLC - draft-ietf-ntp-ntpv5-requirements
>   Ø  I would go further and question why this document needs to be advanced
> beyond the working group at all.
>  I strongly support this.
>  Another advantage is that RFCs never change, whereas we can always just post a new draft when we want to make changes.
>  I believe other WGs have done this kind of thing.
> _______________________________________________
> ntp mailing list
> ntp@ietf.org
> https://www.ietf.org/mailman/listinfo/ntp
> _______________________________________________
> ntp mailing list
> ntp@ietf.org
> https://www.ietf.org/mailman/listinfo/ntp