Bill Weinman wrote:
I have just finished a new draft of the AMTP specification. I would
welcome comments from this group.
I've submitted it to the I-D editor, but until they get to it you'll
find a copy here:
http://amtp.bw.org/docs/draft-weinman-amtp-01.txt
Yakov invited me to join this group after the original draft, but I
found myself overwhelmed with response from slashdot and on the
amtp-discuss list. Anyway, this draft is a lot more complete.
TIA for your comments,
Major nitpick: "roll" is a kind of bread; you mean "role".
Now, for my real beef: the DNS requirements break the use of
'residential' ISP lines for hosting AMTP servers. I currently host an
SMTP server on my local network that would do direct delivery, except
that a certain large network doesn't accept delivery from
dynamic/residential pools. So, I have to use my ISP's slow, unreliable
'smart' host SMTP server as a relay. Your design codifies this
irritating standard as mandated policy on an Internet-wide basis.
Even if I used my ISP-provided rDNS as the argument for EHLO, and
purchased a certificate from an outside CA with the same as the CN, my
ISP can change this value at any time, thus invalidating my certificate
and making mail delivery impossible.
You do address this, but with the assumption that the ISP is willing and
able to be helpful. This is not a valid assumption in many circumstances.
Philip Miller
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg