On Tue, Feb 4, 2020 at 12:25 PM Dotzero <[email protected]> wrote:

> I am not against experiments, but having reread the entire thread starting
> from Dave's post in August, I believe his concerns are valid.
>

I want to say again that I make no assertion that any of Dave's claims are
invalid.  The main thing I want us to discuss is whether any of them rise
to the level of halting PSD's movement through the process versus finding a
way to do all of it in parallel, accepting the risks being identified.
Fixing everything first will be extremely time-consuming, but it is a
possible course of action.

My question to the chairs and the group as a whole is whether an experiment
> can be constructed that is valid and useful without "comingling" PSD issues
> and concerns with the core of DMARC at scale? That is, the group that is
> seriously interested does their experiment amongst themselves to produce
> data that supports and justifies such changes in the wild.
>

Yes, that is the question.  And the working group can still legitimately
conclude that it wants to advance this document before conducting that
experiment.  As you said, rough consensus.

I would remind everyone, however, that this document still needs to pass a
two week IETF-wide Last Call, appropriate directorate reviews, and IESG
review before it can be sent to the RFC editor queue, where again it will
wait for a while.  Assuming we have consensus to proceed in spite of what
Dave and now Mike are saying, we're still at least a month or two from
publication.  That seems a long time to wait before starting the experiment
and collecting data.  Are we sure we want to serialize everything this way?

-MSK
_______________________________________________
dmarc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmarc

Reply via email to