On Wed, Oct 14, 2026, at 11:43, William Weiner wrote: > Hannah, > > The concern is not what the spec says. It is what implementations build from > chain depth data once it is available. Spam filters and reputation systems > use available signals regardless of whether a BCP tells them to. A chain with > depth 1 is a different feature value than a chain with depth n>1, and systems > will train on that difference. The guidance exists so they do not infer more > meaning than is intended. This is to address human factors concern.
A huge chunk of email goes direct. A chain with depth == 1 will be the common case. I believe (and I think Hannah does too) that you are chasing a shadow here - there is no substance in this worry. Bron. -- Bron Gondwana, CEO, Fastmail Pty Ltd / Fastmail US LLC [email protected]
_______________________________________________ Ietf-dkim mailing list -- [email protected] To unsubscribe send an email to [email protected]
