Re: misconception in SPF

On 7 Dec 2012, at 16:32, SM wrote:

> I don't recall whether noreply@example is used in the envelope for 
> delivery failures (assuming valid messages).

If that sentence is not mis-worded, it is a declaration of ignorance of 
important fundamental facts about how email works. RFC5321 is required 
reading for anyone who wants to be taken seriously in technical 
discussions of email.

> There are valid messages with an invalid Return-Path.

This could only be true when using a technically useless definition of 
"valid messages" which does not refer to any formal standard or even to 
informal best practice norms.

There are many strong arguments for mail systems to refrain from 
attempting to require strict compliance to many details of the SMTP and 
email format standards, but it is unhelpful to assume that historically 
or even currently widespread forbearance of operational enforcement of 
some facet of the standards grants any sort of validity to non-compliant 
messages.

> This may due to some assumptions about how email works.

Indeed.