Unsolicited ads

Unsolicited ads

[David Worenklein]

Some how, I've gotten onto an e-junk-mail list.  Has anyone thought
about ways of dealing with this?  For instance, maintaining a list of
all people who have sent me mail in the past and lowering the score of
an article if it's from someone not on the list?

Re: Unsolicited ads

[Per Abrahamsen]

>>>>> "DW" == David Worenklein <dcw@gcm.com> writes:

DW> Has anyone thought
DW> about ways of dealing with this?  

Complain.  I usually forward a copy to `postmaster', `abuse', and the
technical and administrative contact (from whois) of the first host(s)
mentioned in the `received:' headers and the DNS server of that host.

It is a boring and time consuming job, but I think it is the only way
to keep the noise at a manageable level.

Re: Unsolicited ads

[Ketil Z Malde]

David Worenklein <dcw@gcm.com> writes:

> Some how, I've gotten onto an e-junk-mail list.  Has anyone thought
> about ways of dealing with this?  For instance, maintaining a list of
> all people who have sent me mail in the past and lowering the score of
> an article if it's from someone not on the list?

I maintain a list of all domains that have repeatedly sent me spam in
the past, and use procmail to automatically discard mail from these
domains.  Some magic ensures that messages that are replies to my own
messages won't be affected.

Of course, the list of rotten domains should perhaps be globally
maintained, so that when enough spams have originated from some domain,
it could be automatically blocked by all who wanted.

For my particular spam filtering, look at:
http://www.ii.uib.no/~ketil/mailsorting.html 

~kzm

Re: Unsolicited ads

[Steven L Baur]

>>>>> "Ketil" == Ketil Z Malde <ketil@ii.uib.no> writes:

>> David Worenklein <dcw@gcm.com> writes:
>> Some how, I've gotten onto an e-junk-mail list.  Has anyone thought
>> about ways of dealing with this?  For instance, maintaining a list
>> of all people who have sent me mail in the past and lowering the
>> score of an article if it's from someone not on the list?

The problem with this is that the worse offenders use throw-away
accounts, so the first message from that mailbox may well be the
last ...

Ketil> I maintain a list of all domains that have repeatedly sent me
Ketil> spam in the past, and use procmail to automatically discard
Ketil> mail from these domains.  Some magic ensures that messages that
Ketil> are replies to my own messages won't be affected.

Ketil> Of course, the list of rotten domains should perhaps be
Ketil> globally maintained, so that when enough spams have originated
Ketil> from some domain, it could be automatically blocked by all who
Ketil> wanted.

Yup.  There is a maintained black list.  Brian Edmonds
<edmonds@cs.ubc.ca> has a reference to it in a Gnus/antispam page:
	http://www.cs.ubc.ca/spider/edmonds/usenet/gnus/
(there's a reference to this in the Gnus FAQ).

Ketil> For my particular spam filtering, look at:
Ketil> http://www.ii.uib.no/~ketil/mailsorting.html 
(Also referenced by the Gnus FAQ, BTW).

The problem with any solution of this form is that it is reactive, and
ineffective against many spams.  A better long-term solution is going
to be maintaining a ``white list'', and filtering non-white list
e-mail to a lower priority mailbox.

I like the idea of only accepting personal e-mail that contains ecash
postage of some sort, and then only cashing in on the postage if you
deem the mail message not worth the time to read it.  The technology
for implementing this kind of a solution is well within the current
capabilities of Gnus all by itself, though it won't be of much use
until ecash (of whatever form) is more prevalent.

A wild idea off the top of my head:
How effective would a closed mailing list be that distributed the
equivalent of NoCEM information about spammers?  The messages would be
of the same form, and could be automatically processed by Gnus ...

-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be proofread for $250/hour.
Andrea Seastrand: For your vote on the Telecom bill, I will vote for anyone
except you in November.

Re: Unsolicited ads

[Steven L Baur]

>>>>> "Per" == Per Abrahamsen <abraham@dina.kvl.dk> writes:

>>>>> "DW" == David Worenklein <dcw@gcm.com> writes:
DW> Has anyone thought
DW> about ways of dealing with this?  

Per> Complain.  I usually forward a copy to `postmaster', `abuse', and
Per> the technical and administrative contact (from whois) of the
Per> first host(s) mentioned in the `received:' headers and the DNS
Per> server of that host.

When you complain be firm but polite when directing the email to the
domain contacts.  In some cases I've had to deal with, the technical &
administrative contacts for spamming domains are lower-level staff who
don't really have much control over the matter.

Don't expect an answer, and resist the temptation to email them an
emacs core dump, as it will probably hurt your system more when it
bounces.

You can always try the tactic of billing (via snail mail) them for your
time in cleaning up after their spam.  This is reasonably effective
against repeat spamming all by itself, however there are enough people
without a clue and outside of the local jurisdiction that you probably
can't sue when they don't pay, and do it more than once (like the
Internet Business Bureau, which I see has changed its ``good'' name in
the last month -- at least Internet Business Broadcasting is a more
appropriate name for a Spammer).

Per> It is a boring and time consuming job, but I think it is the only
Per> way to keep the noise at a manageable level.

Yup.
-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be proofread for $250/hour.
Andrea Seastrand: For your vote on the Telecom bill, I will vote for anyone
except you in November.

Re: Unsolicited ads

[Lars Magne Ingebrigtsen]

Steven L Baur <steve@miranova.com> writes:

> A wild idea off the top of my head:
> How effective would a closed mailing list be that distributed the
> equivalent of NoCEM information about spammers?  The messages would be
> of the same form, and could be automatically processed by Gnus ...

I think it sounds like a good idea.  It would have to react to email
addresses and not Message-IDs, though.  But a mailing list that one
could subscribe to that will instruct your mail reader which persons
to ignore would be a nice idea.  Properly PGP-signed and all that
jazz, of course.  

-- 
  "Yes.  The journey through the human heart 
     would have to wait until some other time."

Re: Unsolicited ads

[Steven L Baur]

>>>>> "Lars" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

Lars> Steven L Baur <steve@miranova.com> writes:
>> A wild idea off the top of my head:
>> How effective would a closed mailing list be that distributed the
>> equivalent of NoCEM information about spammers?  The messages would be
>> of the same form, and could be automatically processed by Gnus ...

Lars> I think it sounds like a good idea.  It would have to react to email
Lars> addresses and not Message-IDs, though.  But a mailing list that one
Lars> could subscribe to that will instruct your mail reader which persons
Lars> to ignore would be a nice idea.  Properly PGP-signed and all that
Lars> jazz, of course.  

Of course.  With the ability to 
1.  choose who you want to accept Mail NoCEM messages from
2.  have the ability to sort away killed messages into specific group
	(mail.edmonds.killed, eg.)
3.  protect certain groups (just in case)
4.  ignore any local addresses (as per the boss-mail example in the manual)
5.  ignore any addresses found on a white list (Hey, that's my
    former roommate in your global kill file!)

You would want to keep the terms simple though, let GroupLens handle
more complex scoring ...
-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be proofread for $250/hour.
Andrea Seastrand: For your vote on the Telecom bill, I will vote for anyone
except you in November.

Re: Unsolicited ads

[Lars Magne Ingebrigtsen]

Steven L Baur <steve@miranova.com> writes:

> Of course.  With the ability to 
> 1.  choose who you want to accept Mail NoCEM messages from
> 2.  have the ability to sort away killed messages into specific group
> 	(mail.edmonds.killed, eg.)
> 3.  protect certain groups (just in case)
> 4.  ignore any local addresses (as per the boss-mail example in the manual)
> 5.  ignore any addresses found on a white list (Hey, that's my
>     former roommate in your global kill file!)

Yup.  But the most important thing is to define the syntax of these
files.  The rest is just a matter of user interface.

Anybody up for writing a standard?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Ingebrigtsen

Clarification (was Re: Unsolicited ads)

[Steven L Baur]

On the 10th of June I wrote in response to:
Lars> I think it sounds like a good idea.  It would have to react to email
Lars> addresses and not Message-IDs, though.  But a mailing list that one
Lars> could subscribe to that will instruct your mail reader which persons
Lars> to ignore would be a nice idea.  Properly PGP-signed and all that
Lars> jazz, of course.  

sb> Of course.  With the ability to 
sb> 1.  choose who you want to accept Mail NoCEM messages from
sb> 2.  have the ability to sort away killed messages into specific group
sb> 	(mail.edmonds.killed, eg.)
         ^^^^ ^^^^^^^ ^^^^^^
...

It is possible to read this the wrong way.  I was not referring to
sorting killed messages based on killee, but rather referring to the
source of the source of the killfile which killed the message.  While
I would be interested in using Global Kill Files coming from a variety
of places, the one which immediately came to mind was Brian Edmonds
Global Kill File stuff quoted in the Gnus FAQ.  I personally would not
trust such a mechanism without some kind of safety net to see what
exactly was being killed by those particular rules.

I offer my public apologies to Brian for any confusion my poor choice of
wording may have brought.

-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be proofread for $250/hour.
Andrea Seastrand: For your vote on the Telecom bill, I will vote for anyone
except you in November.