Re: [EXIM] DynamicIP Spam Sources List (DSSL)

Top Page
Delete this message
Reply to this message
Author: Ronald F. Guilmette (exim-users)
Date:  
To: exim-users
Subject: Re: [EXIM] DynamicIP Spam Sources List (DSSL)

In message <Pine.LNX.3.96.990427014333.20195O-100000@???>,
Jason Gunthorpe <jgg@???> wrote:

>
>On Sun, 25 Apr 1999, Ronald F. Guilmette (exim-users) wrote:
>
>> >> number of hits you got from each of the two lists. I'm confident that
>> >> you will see more through spam stopping when consulting the DSSL list.
>> >
>> >Hits does not nessarily equate to spam blocked.
>>
>> OK, so arrange to have your mail server siphon off everything that comes
>> in that matches against the DSSL for a week (or two). Then, at the end
>> of that time, go through and look at all of the messages that have been
>> caught in your net.
>
>Hmm, you know I have actually done this (in a fashion..) using the IP
>addresses the mailer logs for incoming connections I sorted/uniq'd them
>and ran them across the MAPS dial up user list - the result was that 14%
>of the sites connecting to lists.debian.org...


A non-representative sample, if ever there was one.

>... would have been blocked.


More precisely, then would have been asked to either route their outgoing
E-mails through their own ISP's ``smart host'' mail server(s) (which I
assume is at least as trivial to do with Exim as it is with Sendmail,
right?) or else to ask their ISP for a static IP address for their dial-up
connection.

>We are lucky to get even 1 spam a day.. so those 14% were certainly not
>sending spams.


Perhaps not, but neither do they represent cases/people who would have been
unable to send mail to you if you implemented DSSL or MAPS DUL blocking of
direct SMTP connections.

I am of the opinion that people who send mail to `lists.debian.org' are more
likely (by far) than the average Internet user to be capable of making
small technical adjustments to their outgoing mail routing... probably
without even getting flustered about it.


--
*** Exim information can be found at http://www.exim.org/ ***