--On Monday, 23 April, 2007 23:48 -0400 Tony Hansen
<tony(_at_)att(_dot_)com> wrote:
The following issue with 2821bis-01 is still open, but we have
a tentative conclusion that needs to be affirmed after
2821bis-02 comes out.
1) Format of domain name - does it allow/not-allow/require
period/trailing-period in TLD?
The discussion on this issue was all over the place with no
totally clear consensus. The trend was toward no dot, but
adding some explanations about top level domains.
An email was sent to the ICANN to find out why some country
code TLDs have MX records, but no response has been received.
Not quite true ... sorry for not noting the responses to the
list when they arrived, but I've been trying to avoid adding to
the clutter.
The final action at ICANN was to send a copy of my note out to
the ccTLD list. So far, I've received about a half-dozen
responses. They varied from one domain that was not using
these, had noticed that they often didn't work, and considered
them dumb to someone who has an MX record but wanted to start a
debate about "services" of this variety and maybe making
Verisign support it in .COM so that MTA/MUA vendors would be
forced to support it.
I also received a couple of long messages from a usual suspect
in these matters (and from an organization that may or may not
consistent entirely of him), claiming to advise and/or represent
several ccTLDs, indicating that SMTP was not the right situation
(although he still cares about MX records although I don't
presume he knows what they are) and the DNS isn't either and
trying to put this all in context with the nationally-based
"multilingual Internet". I assume those comments apply to the
well-known X.400(2008) and not to SMTP or RFC2821bis.
Bottom line, based on responses received so far, is that no one
is using mail to TLD names. Some think they would like to.
My impression, based on those remarks but, more important, on my
reading of the comments on the list, is that we shouldn't ban
mail to TLD domains, but should clarify and strengthen the "only
FQDNs on the wire" rule. That impression is instantiated in the
-02 draft although, of course, it could be changed again if it
isn't right.
When 2821bis-02 comes out, please review the text in section
2.3.5.
please
john