In article <232fc698-bdc7-4be1-9e14-da6c788f1...@dukhovni.org> you write: >To make sure that we're talking about the same thing, I want to check >that you're proposing: > > Received: from client.example.com (client.example.com [192.0.2.1]) > by mta.example.net (split-personality-sni.example.net [192.0.2.2]) > ... > >where the "TCP-Info" in the "BY" clause records the SNI name?
Yup. >many decades ago? Starting with Sendmail versions going back to at >least the mid 1980's, the "comment" after the "BY" clause has been >used primarily used to record the MTA software version, though usages >do very. Here are some trace headers from a single message: > > Received: from zardoc.esmtp.org (zardoc.esmtp.org [75.101.48.117]) > (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) > (No client certificate requested) > by mournblade.imrryr.org (Postfix) with ESMTPS id AB5307A3309 Ah, but they're not the same thing. The SNI string goes in the tcp-info part of the extended-domain rule, but your (postfix) is just a comment. There's no IP address after the postfix so you can tell it's not tcp-info. "(zardoc.esmtp.org [75.101.48.117])" is tcp-info since it's the correct syntax. I realize all of this is hairsplitting and a reasonably good header parser can look in all of the places where you might hide something, but when possible I'd prefer to use the syntax we've got rather than faking up something new. R's, John _______________________________________________ Uta mailing list Uta@ietf.org https://www.ietf.org/mailman/listinfo/uta
