[horde] Blacklist and forwarding

[Vilius Sumskas/LNK]

>>> Hi Everybody.
>>> Thanks for the answers.
>>>
>>> The statement from Vilius really makes me change my perception
>about the
>>> blacklisting feature, where I believe is made to block a single
>> mailaddress
>>> like a at a.com. But I don't see another option to make the
>forwarding
>> process
>>> easier to the user.
>>> I believe that my suggestion can work because DSPAM will be
>trained
>> based on
>>> the signature (token) of the header of the message, no matter who
>the
>> sender
>>> is.
>>
>> I understand that, but DSPAM (or any other spam training
>techniques) have
>> nothing to do with blacklisting. So what I'm trying to say is that
>you
>> should NOT try to use blacklist for that.
>>
>>> Is possible to implement that? I mean, when I  right-click an
>e-mail and
>>> select blacklist, the system performs this extra step.
>>
>> As Michael already said - no. But it makes no sense to do this...
>see my
>> above reply.
>>
>>> Is there another option? Like activate the report spam feature
>and
>> attach it
>>> to a forwarding process.
>>
>> Yes exactly IMP has a spam reporting functionality, see IMP's
>> configuration for more details. You can also search Google with
>something
>> like "IMP spam reporting". You just need to specify your program or
>shell
>> script which will be responsible for reporting and that's it. And
>TELL
>> users to choose Mark As Spam instead of Blacklist when they
>right-click
>> email.
>>
>> For example in spamassassin you specify "/usr/bin/spamassassin -r",
>in
>> Razor you can use "/usr/bin/razor-report" for spam and
>> "/usr/bin/razor-revoke" for ham. I'm not sure how DSPAM works but
>I
>> suppose it should have a possibility to report this similary. If
>not, then
>>
>> just create a shell script on your own and make it send email to
>the
>> address you need. It should be like 2 or 3 lines of code.
>>
>> --
>>    Vilius
>
>
>
>Hi
>I've made it! And it was so easy... I have just configured IMP to  
>forward the e-mail to spam at mydomain (when user report as spam) and  
>notspam at mydomain(when user report as ham). It worked!!!!
>
>Check my settings:
>
>'spam' => array(
>          'innocent' => array(
>              'display' => true,
>         //
>         //     // Email reporting driver
>              'email' => 'notspam at mydomain',
>              'email_format' => 'digest',
>         //
>         //     // Null reporting driver
>         //     'null' => true,
>         //
>         //     // Program reporting driver
>         //     'program' => null
>          ),
>          'spam' => array(
>              'display' => true,
>         //
>         //     // Email reporting driver
>              'email' => 'spam at mydomain',
>              'email_format' => 'digest',
>         //
>         //     // Null reporting driver
>         //     'null' => true,
>         //
>         //     // Program reporting driver
>         //     'program' => null
>         // ),
>         // // It is possible to directly define additional spam
>drivers.
>         // // The 'drivers' array should contain a list of driver
>objects
>         // // (these classes must implement the IMP_Spam_Base
>class).
>         // 'drivers' => array(
>         //     new IMP_Example_Spam_Driver()
>          )
>
>With that, DSPAM is receiving the e-mail on my SMTP server (which is
>
>in another host), processing the token and training the database.
>
>I believe that this settings are correct, right?
>
>Thanks a lot for your help!

Didn't know new IMP version could do that. If it's working then settings are correct.

-- 
   Vilius