Hi Heiko,
Thanks very much for this information - so two more questions for you and the community:
1) It seems that ACL is faster when compared to TRANSP?ORT - is this true?
2) Is Exim planning on removing the ability to perform the TRANSPORT type of operation?
On Feb 4, 2012, at 11:14 AM, Heiko Schlittermann wrote:
> Hello Antonio,
>
> Antonio Leding <dev@???> (Sa 04 Feb 2012 04:19:20 CET):
>> In reading the Exim documentation, it is clear that SpamAsassin can be called by Exim at the desired time by using ACL control. Then, it seems that SpamAssassin will hand back a SPAM score & report so that Exim can then take the information and move forward with message processing. With that in mind, here are my questions:
>
> The "desired time" will almost always be the DATA ACL, it's the only
> place, where you have the message to scan. Keep in mind, that at the time of
> the DATA ACL you may have more than one recipient, thus scanning it with per
> recipient rules can't be done here. (But - you can limit the number of
> recipients per message in the RCPT ACL, this would give you the
> possibility of implementing something that is about a per recipient rule
> (ignoring things as aliases and so on.))
>
> Other modes of interaction with SA are the system filter and the
> transports filter. The latter one will be called once per transport, and
> since normally a transport is called for each recipient, you may
> implement per recipient spamassassin rules.
>
> Using SA in the ACL is using it as an "spam evaluator", just *returning*
> you the spam status, spam report, … of the message. Using SA as a filter
> is using it really as a *filter*, the message passes SA completly. This
> filter mode allows SA to modify the message (rewrite the subject,
> rewrite the body). If you use SA as an evaluator (in the DATA ACL), the
> modification of the subject line (and possibly insertion of additional
> header lines) is completly up to Exim.
>
> I'd prefer the DATA ACL, but you may have reasons not to do.
>
>> 1) Are Exim ACL and Transport modes mutually exclusive or can they work in tandem?
>
> I wouldn't say that it's mutually exclusive, but probably you're doing
> the scan twice.
>
> --
> Heiko :: dresden : linux : SCHLITTERMANN.de
> GPG Key 48D0359B : 3061 CFBF 2D88 F034 E8D2 7E92 EE4E AC98 48D0 359B
> --
> ## List details at https://lists.exim.org/mailman/listinfo/exim-users
> ## Exim details at http://www.exim.org/
> ## Please use the Wiki with this list - http://wiki.exim.org/
