On Fri, Jul 10, 2020 at 1:40 AM Sara Dickinson <s...@sinodun.com> wrote:
> > I suggest getting rid of use of BCP 14 entirely. There are only two > SHOULDs in > > the whole thing, and I don't think you need them. > > This point has been discussed a few times - the WG considered a few > alternatives and this was what eventually got consensus. We also added new > text in the -12 version (suggested by Ben Kudak) at the end of section 5 to > clarify the point that there are normative requirements here: “The rest of > this document does not use normative language but instead refers only to > the three differing classes of action which correspond to the three named > levels of compliance stated above. However, compliance (to the indicated > level) remains a normative requirement.” If you want to suggest a further > update to this text, please do. > Hi Sara, thanks for your consideration. As far as I can tell -- and I fully admit that's without the benefit of having been part of the WG's deliberations -- the two SHOULDs ought to be MUSTs, otherwise an operator could do neither of them and still be compliant with the BCP because "I have my reasons". And if you're going to make them MUSTs, then you can just say "implements" instead of "SHOULD implement" and "publishes" instead of "SHOULD publish", and then you don't need BCP 14 at all. The normative force of the BCP's text is not reduced merely by not using BCP 14. All that said, this is a comment to a YES ballot, so my advice is worth what you paid for it. :-) -MSK
_______________________________________________ dns-privacy mailing list dns-privacy@ietf.org https://www.ietf.org/mailman/listinfo/dns-privacy
