Hi Éric,
Thank you for your review!
On 5/20/26 09:18, Éric Vyncke via Datatracker wrote:
## COMMENTS (non-blocking)
### Andy's DISCUSS
I second Andy Newton's DISCUSS point about the use of `SHOULD` vs. `MUST`. The
sentence in section 3 is not enough IMHO `However, not following any
requirements designated with the "SHOULD" key word will generally lead to
undesirable effects of ambiguity and interoperability issues.`
Andy has proposed improved wording:
https://mailarchive.ietf.org/arch/msg/dnsop/X2bmBcl8Bx6sBRts8fc_U5mKxdE/ Unless
I hear back, I'll assume you're OK with that.
The use of BCP14 is also questionable as it is not really about interoperation.
E.g., section 4.2.2 contains `The reduction should be in effect at least for a
couple of days` still providing guidance to the operators without the use of
BCP14. Same for section 5.2, which has no BCP14 and still useful.
The reason for this is as follows: Sections *.1 are the recommendations
sections and sections *.2 contain analysis and rationale. Sections *.1 are
compiled verbatim into Appendix A so you can find all the recommendations
relevant for DS automation interoperability by just look at the Appendix, as a
handy implementation reference.
The recommendations sections therefore contain the distilled core of the recommendations.
The particular case you quote ("The reduction should be in effect at least for a
couple of days") is a very qualitative statement, and in my view not an
interoperability statement that would really add to what the corresponding recommendation
4.1.2 says; it rather provides additional context.
In case this doesn't clarify the issue, feel free to re-raise and we can see
together what to do about it.
See also next issue.
### BCP status
Linked to Andy's DISCUSS, whether this document is BCP or 'informational' is
also another question. Some BCP-like RFC (e.g., RFC 9099) are informational and
do not rely on BCP14, only plain English "should".
I don't see Andy raising a question about the document status.
Roman was concerned; we've attempted to address that (pending his feedback) by
ensuring that defining recommendations are MUST, not SHOULD (which indeed was
an error).
For extra context on the BCP status, please see the shepherd write-up.
Thanks,
Peter
_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]
