reporting known/unknown recipients

Matus UHLAR - fantomas uhlar@fantomas.sk
Thu Apr 15 14:32:29 UTC 2010


> > From: Matus UHLAR - fantomas <uhlar@fantomas.sk>
> 
> > The thing about DCC is, that I've seen it reject some valid "chain" mails
> > (jokes etc) that aren't apparently quite correct and apparently not abusive
> > in any way. I don't want to reject purely on DCC threshold.

On 15.04.10 13:50, Vernon Schryver wrote:
> Spam is unsolicited bulk email, but DCC only detects "bulk."

I'm sure that every mail to nonexisting address is unsolicited :)
(well, nearly... sometimes mail is sent after address was disabled)

> So one needs to combine DCC with a mechanism for detecting "solicited."

If there was mechanism comparing number of reports to number of mails
received, my proposal would have the opposite result - it would lower the
reports/total rate. But since there's none, I had the idea if we can push
the count towards the expected result.

Of course, I'm asking and waiting for arguments.

> Whether a given copy of a bulk mailing is solicited depends on the
> individual target, so the best way to detect "solicited" is with
> per-user, individual whitelists.  Something like the cgi scripts
> in the DCC source can be used (and are at some installations) to 
> let users handle their own whitelists.  Those scripts are demonstrated
> at http://cgi-demo:cgi-demo@www.rhyolite.com/dcc-demo-cgi-bin/
> Many other DCC installations use a scoring system such as SpamAssassin.

I'm not discussing the ways to measure whether mail is spam, but ways to
measure whether it's bulk. My proposal was to count mail to one non-existent
address as two or more mails to one existent address.

-- 
Matus UHLAR - fantomas, uhlar@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
Quantum mechanics: The dreams stuff is made of. 



More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.