spf-discuss
[Top] [All Lists]

Re: Re: draft-schlitt-spf-02 now available and submitted to the IETF

2004-12-28 18:05:34
On Tue, Dec 28, 2004 at 07:26:53PM +0000, Leonard Mills wrote:

Hosts do not identify themselve as domains! Hosts can use domain as part 
of the host's FQDN though.. So change wording in some fashion, I suggest:

Hate to be pedantic, but even back in 82[12] days the email RFCs 
identified email addresses as (meaning, not the literal BNF):

   <local-part>@<domain>

Thus, the HELO domain can be indistinguishable from a fully
qualified domain name.  But it _is_ the domain for email purposes.

I think you're missing the point as well (although not so bad):

If the hostname is "somehost.someprovider.sometld" then:

a) somehost.someprovider.sometld is a domain
b) someprovider.sometld is a domain
c) sometld is a domain

and no, this is _not_ a multiple choice question.

"user(_at_)somehost(_dot_)someprovider(_dot_)sometld" -> <local-part>@<domain>
"user(_at_)someprovider(_dot_)sometld"          -> <local-part>@<domain>

is valid in both cases.  These domains differ but they both
qualify as domains.  It is not just the domain for email
purposes; both are domains, period.

For HELO purposes, the domain to use is the FQDN of the
sending host so there's no choice but to use
"somehost.someprovider.sometld" (except for the address-literal
case of course).

This may seem nitpicking but since we're going to define
a standard, it is important to get it right.  If we can't
even get DNS right, how are we going to get SPF right when
it builds on DNS ?!?

Alex


<Prev in Thread] Current Thread [Next in Thread>