X-ASVP is a public specification.
I see an expired Internet Draft from last year. Is there anything more
recent?
If one were to use the X-ASVP framework to announce the use of VBR-Info,
then instead of (or in addition to) hosting a "token generator" at their
TGURI, they would post :
X-ASVP:V9[SUPPORTED,VBR-Info]
Recipients can tell that senders use VBR because there is a VBR-Info:
header, just like they can tell that senders use DKIM from the
DKIM-Signature: header, or that they use S/MIME from Content-Type:
multipart/signed. What would be the advantage of a second header
announcing the presence of a first header?
I don't currently understand from the VBR-info specification how a
sender is expected to know that the recipient uses the method to
filter their messages.
They don't. They put the VBR-Info header on all their mail, just like
they probably put DKIM-Signature as well. The incremental cost of
another header is insignificant. We designed VBR to minimize the
number of DNS lookups that recipients will do on incoming mail,
typically no more than one or two per VBR message.
X-ASVP is intended to allow the advertisement of multiple methods.
So a recipient may advertise ...
I gather from your draft that senders are expected to do a per-message
HTTP lookup to see what the recipient supports. To a rough
approximation that second TCP session doubles the sender's cost of
mail delivery. Why would a sender want to do that rather than just
put whatever they use on all their mail?
Also, SMTP is store-and-forward and signature based systems like DKIM
and S/MIME are designed to work through multiple hops. Are
intermediate systems supposed to do call-forwards to find out what
later systems will use, or what?
R's,
John
_______________________________________________
Asrg mailing list
Asrg(_at_)irtf(_dot_)org
https://www.irtf.org/mailman/listinfo/asrg