Re: [dns-privacy] [Ext] WGLC : draft-ietf-dprive-unilateral-probing

[Rob Sayre <[e-mail]>]

On Fri, Jun 9, 2023 at 3:44 PM Hollenbeck, Scott <[e-mail]>
wrote:

> *[SAH] The IESG deliberately chartered this working group to “Investigate
> potential solutions for adding confidentiality to DNS exchanges involving
> authoritative servers” in an Experimental manner. As Brian noted, that’s a
> binding agreement with the IESG. We can either do that or attempt to
> re-charter the working group. I’m under the impression that Brian’s last
> note to the group was a request to discuss those two options, which could
> include discussion of how to conduct the experiment. It’s not an ad-hoc
> process at all.*
>

Hi,

I agree that a recharter would be required. However, what you're asking for
here exceeds the requirements of a Proposed Standard, so that does seem a
bit ad-hoc to me.

https://www.rfc-editor.org/rfc/rfc7127.html#section-3.1

In particular this paragraph applies:
"The IESG may require implementation and/or operational experience
prior to granting Proposed Standard status to a specification that
materially affects the core Internet protocols or that specifies
behavior that may have significant operational impact on the
Internet."


> I never like to read stuff like this. Each of us probably has a regulator
> that annoys us in their treatment of some issue. But we can't really make
> decisions based on guesses about the future actions of unnamed regulators.
> I'm also sure you know the document ladder quite well, but you've used
> imprecise terms here. In the first sentence, you say "IETF standards". But
> the last sentence says "proposed standard".
>
>
>
>
> *[SAH] I used those terms deliberately. My employer has contractual
> obligations to implement a mix of IETF-developed Proposed Standard and
> Standard specifications – that is, “IETF standards”. In the last sentence,
> “proposed standard” specifically refers to one possible status for this
> draft.*
>

So, your employer has contractual obligations to implement some
IETF-standards track documents. I'm still a little mystified, because I
don't think anyone would sign or write such an agreement for documents
not-yet-written. I figured the objection would be the typical
encryption-related ones (cost, observability, etc). The sort of thing we
saw with HTTP2 and DoT/DoH/DoQ.

But I also originally wrote that Experimental would work here, even if the
label is inaccurate.

thanks,
Rob