ALL_SPAM_TO not working correctly?
I have run across the following situation: I have a user, which receives all spam unmodified (ALL_SPAM_TO). When a spam message is sent to multiple users on my machine, including the one in ALL_SPAM_TO, all users addressed in the message get it unmodified, not only the ALL_SPAM_TO user. Is this correct behaviour? -Sietse
Re: ALL_SPAM_TO not working correctly?
Sietse van Zanen wrote: I have run across the following situation: I have a user, which receives all spam unmodified (ALL_SPAM_TO). When a spam message is sent to multiple users on my machine, including the one in ALL_SPAM_TO, all users addressed in the message get it unmodified, not only the ALL_SPAM_TO user. Is this correct behaviour? -Sietse SA doesn't know for sure who the current message is being delivered to. It acts only on the contents of the message, nothing more. To compound the problem, if you call at the MTA layer, there is only one message fed to SA. At that point, SA absolutely must act on an all or nothing basis. If you're calling at the MDA layer in a way that allows per-user user_prefs files, move the all_spam_to command into that user's own user_prefs file.. This way it will only be in effect when the message is being delivered to that user.
RE: ALL_SPAM_TO not working correctly?
I figured it would be something like that. I have moved the spamsink to the milter config. The milter should replace all recipients with only the spamsink. -Sietse From: Matt Kettler Sent: Fri 08-Dec-06 13:13 To: Sietse van Zanen Cc: users@spamassassin.apache.org Subject: Re: ALL_SPAM_TO not working correctly? Sietse van Zanen wrote: I have run across the following situation: I have a user, which receives all spam unmodified (ALL_SPAM_TO). When a spam message is sent to multiple users on my machine, including the one in ALL_SPAM_TO, all users addressed in the message get it unmodified, not only the ALL_SPAM_TO user. Is this correct behaviour? -Sietse SA doesn't know for sure who the current message is being delivered to. It acts only on the contents of the message, nothing more. To compound the problem, if you call at the MTA layer, there is only one message fed to SA. At that point, SA absolutely must act on an all or nothing basis. If you're calling at the MDA layer in a way that allows per-user user_prefs files, move the all_spam_to command into that user's own user_prefs file.. This way it will only be in effect when the message is being delivered to that user.