4

I always thought that the return-path header was used for bounced messages, so I have my email newsletter software setup to unsubscribe anyone whose email address replies back to the return-path email address.

But recently some users started telling me their out-of-office replies where unsubscribing them. Do out-of-office replies get sent to the return-path on some email servers?

I could check all emails to that address and see if they have the "message/delivery-status" content-type in one of the parts, but I wasn't sure if this was necessary.

Brian Armstrong
  • 1,557
  • 3
  • 18
  • 22

3 Answers3

5

Out-Of-Office replies do usually get sent via return-path, yes.

It's correct to the RFC specs, too. More info here: https://www.rfc-editor.org/rfc/rfc3834

Community
  • 1
Chris Thorpe
  • 9,903
  • 22
  • 32
  • Great info...thanks! So I'm guessing at this point the most reliable way to detect bounces to look for the "message/delivery-status" content-type in emails to that address - or possibly parsing postfix logs. I'm surprised there isn't a more well defined standard. Thanks for the help! – Brian Armstrong Mar 30 '10 at 23:49
3

I have been doing a lot of research on this recently. Exchange out of office messages seem to use the from: and not the returnpath: when sending out of office messages. This seems to be improper behavior, because the from is a friendly display version of the returnpath, but the returnpath is what replies should be sent to.

For your newsletter system, out of office messages from exchange will probably not use the returnpath and therefore not unsubscribe. However, you cannot depend on this behavior. Other email systems use the returnpath.

Lookup bounce address tag validation (BATV). the BATV tag is on the returnpath, but not the from. so that you see an email from john.doe@domain.com, but when you reply it goes to the returnpath at prvs=086421=john.doe@domain.com

Paul
  • 31
  • 1
-2

It uses the From address or the Reply-to if set. Return-Path is really for servers not clients.

John Kloian
  • 142
  • 3