Re: X-Auto-Response-Suppress header
<kd6lvw <at> yahoo.com>
2012-05-22 01:28:38 GMT
--- On Mon, 5/21/12, George Roberts <groberts <at> purity.net> wrote:
> I'm sort of sorry I started this
> whole thing, LOL. Just so I have some clarity on the issue,
> could someone please explain to me what exactly it is that
> Exchange does with Received: headers that is so bad? I
> see Received: lines from various mail systems and they all
> seem to me to be a little bit different. Also, if possible,
> please state which RFC dictates the format Exchange is
> accused of breaking.
> Just trying to understand what the issue is.
RFC 821 and its successors and add-ons through to 5321.
It places a multi-word phrase after "with", while "with" takes a single word called an "atom" in the syntax.
Sometimes, it also generates "id" field parameters which contain periods when they are not bracketed
with greater-than/less-than and contain an " <at> " (i.e. an "addr-spec"). Basically, they don't follow the
required syntax and are therefore in error.
If exchange didn't use SMTP, there would be no problem with its "Received:" headers since the stricter
syntax of RFC 5321 wouldn't apply, and these do meet the weak syntax of RFC 5322 (which applies to NON-SMTP messages).
When "with" is present, its parameter must be on the list of IANA registered values. When "with SMTP" (or a
variant) is present, RFC 5321 governs the REQUIRED syntax of the "Received:" header (which includes
mandatory "from" and "by" clauses, etc., ...). "with Microsoft SMTPSVC" simply isn't valid.
Exchange uses SMTP but generates a syntactically incorrect header. Similarly with Google's gmail (it
often omits the "from" clause when required), Yahoo's use of an unregistered protocol ("with NNFMP"*),
qmail, and of late, exim.