You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Charles Gregory <cg...@hwcn.org> on 2010/01/08 20:39:29 UTC

Re: Broken mass mailings - Delivered-To: issue

Please reply to list. This reply is to list...

On Fri, 8 Jan 2010, Richard B. Emerson wrote:
: Actually, the example note has To: baba-l@yahoogroups.com not To:
: chris@pinefields.com...

The To header is irrelevant. Mail to this list is addressed "To:" the 
list. But the *envelope* (which is captured in your system as a 
header "Delivered-To:" is the real recipient.

: ... The intent of the recipe below is to take a message containing 
: Delivered-To: and send it to the addressee in that header entry.

The 'Delivered-To' means it is ALREADY being delivered to that address!
Your MTA (sendmail) should have invoked procmail on behalf of the 
'Delivered-To' user. And this should *not* depend on the "To:" header in 
any way....

: "if X-Loop:..." is present, don't forward the It should keep a 
: forwarding loop from forming.  Why this isn't the case is part of my 
: problem.

Unless the forwarding loop is occurring *outside* of procmail!

: Why am I using sendmail?  Because that's the way it was set up with SuSE
: Linux 11.1, which is what I'm running here. 

No, no, why are you SENDING the mail you just received, rather than 
delivering it directly to the recipient's mailbox? IE:

:0
*(conditions)
$USER/inbox

: Again, keep in mind that however strange things may have been here (with 
: sendmail and procmail), until very recently, it all worked as it should.  

Any time I encounter *this* condition, I always ask, "what changed?" 
Look in your logs for *anything* being updated around that time.

Remember it is the recipient's shell invoked for things like 'formail', so 
even if you changed permissions or the 'path' for a user shell, it could 
have an impact... ANYTHING that changed....

- C

Re: Broken mass mailings - Delivered-To: issue

Posted by Charles Gregory <cg...@hwcn.org>.
Ooops. Sorry. A poster replied offlist, and when I attempted to put
it back on-list, I lamely inserted the SA users list address instead of 
the procmail list address. (smack forehead)

Disregard.....

-C

On Fri, 8 Jan 2010, Charles Gregory wrote:
: Please reply to list. This reply is to list...
: 
: On Fri, 8 Jan 2010, Richard B. Emerson wrote:
: : Actually, the example note has To: baba-l@yahoogroups.com not To:
: : chris@pinefields.com...
: 
: The To header is irrelevant. Mail to this list is addressed "To:" the 
: list. But the *envelope* (which is captured in your system as a 
: header "Delivered-To:" is the real recipient.
: 
: : ... The intent of the recipe below is to take a message containing 
: : Delivered-To: and send it to the addressee in that header entry.
: 
: The 'Delivered-To' means it is ALREADY being delivered to that address!
: Your MTA (sendmail) should have invoked procmail on behalf of the 
: 'Delivered-To' user. And this should *not* depend on the "To:" header in 
: any way....
: 
: : "if X-Loop:..." is present, don't forward the It should keep a 
: : forwarding loop from forming.  Why this isn't the case is part of my 
: : problem.
: 
: Unless the forwarding loop is occurring *outside* of procmail!
: 
: : Why am I using sendmail?  Because that's the way it was set up with SuSE
: : Linux 11.1, which is what I'm running here. 
: 
: No, no, why are you SENDING the mail you just received, rather than 
: delivering it directly to the recipient's mailbox? IE:
: 
: :0
: *(conditions)
: $USER/inbox
: 
: : Again, keep in mind that however strange things may have been here (with 
: : sendmail and procmail), until very recently, it all worked as it should.  
: 
: Any time I encounter *this* condition, I always ask, "what changed?" 
: Look in your logs for *anything* being updated around that time.
: 
: Remember it is the recipient's shell invoked for things like 'formail', so 
: even if you changed permissions or the 'path' for a user shell, it could 
: have an impact... ANYTHING that changed....
: 
: - C
: