[Mimedefang] How can mimedefang use spamc/spamdfuncitonsofspamassassin?

[Jan Pieter Cornet]

On Thu, Apr 14, 2005 at 09:21:50AM -0700, Matthew.van.Eerde at hbinc.com wrote:
> Jan Pieter Cornet wrote:
> > You cannot have per-user configurations, and reject spam based on
> > spamassassin score, in a general way, because of limitations of the
> > SMTP protocol. There is little that mimedefang can do about that.
> > 
> > I am trying to do the same thing though :) but I don't think per-user
> > scoring with the granularity that spamassassin offers, is scalable.
> > I'm aiming for a few classes of SA settings, and probably also
> > dividing remote hosts in spammishness-levels, and then selectively
> > tempfailing multiple recipients when you receive different "classed"
> > users from spammish remote hosts. But that's currently only in the
> > "idea" stage... 
> 
> What is the distribution of scalar @Recipients across your incoming mail?

Just for kicks, this is our distribution for yesterday (Mon April 18th)

1: 635464
2: 54999
3: 29229
4: 24131
5: 8599
6: 10558
7: 7591
8: 5057
9: 2428
10: 1856
11: 1149
12: 1033
13: 830
14: 690
15: 460
16: 307
17: 271
18: 206
19: 272
20: 84
21: 68
22: 69
23: 28
24: 26
25: 22
26: 9
27: 17
28: 13
29: 14
30: 11
31: 19
32: 10
33: 18
34: 10
35: 6
36: 9
37: 5
38: 7
39: 8
40: 9
41: 5
42: 14
43: 14
44: 22
45: 32
46: 28
47: 20
48: 23
49: 
50: 1

So, yes, only 20% of the mails have multiple recipients, and only 12% have 3
or more recipients, but still... it's not a fraction I can easily ignore,
certainly not with these numbers.

I'd like to err on the side of caution... if a mail is suspected to be spam,
I want to either put it in a separate spam folder, or reject it with an
appropriate error message. After running spamassassin on some 300k+ email
addresses, most of whom are "average Joes", it is clear that there is a lot
of wanted email out there that makes spamassassin jump skyhigh (well, upto
9, at most, I haven't yet seen reports of false positives with scores over 
10).

-- 
#!perl -wpl # mmfppfmpmmpp mmpffm <pmmppfmfpppppfmmmf at fpffmm4mmmpmfpmf.ppppmf>
$p=3-2*/[^\W\dmpf_]/i;s.[a-z]{$p}.vec($f=join('',$p-1?chr(sub{$_[0]*9+$_[1]*3+
$_[2]}->(map{/p|f/i+/f/i}split//,$&)+97):qw(m p f)[map{((ord$&)%32-1)/$_%3}(9,
3,1)]),5,1)='`'lt$&;$f.eig;                                # Jan-Pieter Cornet