spf-discuss
[Top] [All Lists]

Re: Re: envelope from vs data from

2004-03-17 09:09:02
----- Original Message ----- 
From: "Meng Weng Wong" <mengwong(_at_)dumbo(_dot_)pobox(_dot_)com>
To: <spf-discuss(_at_)v2(_dot_)listbox(_dot_)com>
Sent: Wednesday, March 17, 2004 9:32 AM
Subject: [spf-discuss] Re: envelope from vs data from


| posting on behalf ...
| 
| On Wed, Mar 17, 2004 at 09:22:01AM -0500, Olson, Margaret wrote:
| | 
| | I like to use "return address" or "envelope from" and "from", because there
| | is a close analogy to postal mail. If you look at a piece of personal postal
| | mail, the return address is very likely to be the same as the person who
| | composed the letter and sent it to you. If they differ, it's mostly likely
| | that the return address is in some sense a subset of the sender's address;
| | it may contain just the household name ("olson") or even just the address,
| | but in any case if it is returned the return address will get the mail back
| | to originator.
| | 
| | If you look at a bill or other piece of non-personal mail, be it a bill, an
| | advertisement, or a begging letter from a charity, it's highly likely that
| | the return address and the from are different - for all kinds of reasons
| | that pretty much apply to bulk email as well. The organization that sent the
| | mail and handles delivery problems is not the same as the one that composed
| | and originated the message.
| | 

I'm sorry, but I'm confused as to the objective to be achieved.

| | I like to use "return address" or "envelope from" and "from", 

If the "envelope from" and "from" are different, does that imply running
the query twice and combining the score based on the result in a
weighted fashion? Maybe pick the one that agrees with the sending 
server?

For example, when evaluating mail from this list, using the "from"
produces a FAIL for anybody having an SPF DNS record and sending
to the list, while using the "envelope-from" or return-path produces
a PASS.

Regards
Greg




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