Message Board

Newbie/Basic Questions

Hony pots and checksum blockers

Author: B.White3 (25 Sep 07 7:06am)

I note the thread below about Pyzor, Razor and DCC.

My current plan was (and still is, though subject to change) to set up a virgin domain, then get some email addresses published, with human readable text making it clear that the emails are for harvesting, and then wait for spammers to spam the email addresses of the domain.

With that I intend on feeding my own Pyzor server.

Mail will not be rejected on the basis of a pyzor hit alone. (Well unless the server has had several copies of the same email) so I see this as a safe way of gathering checksums.

I can donate sub domains, but wondered if I could get a copy of the emails as well?

Also, it would be seriously really handy to feed the checksums to pyzor, razor and DCC servers. They do measure by certainty, and pyzor does support whitelists.

Any comments?

(BTW, I operate 10 spam killing servers, handling millions of connections a week and increasing).

Re: Hony pots and checksum blockers

Author: M.Prince (25 Sep 07 7:01pm)

We don't have any way to provide you copies of the messages your subdomains receive. There are a number of architectural reasons that it's unlikely we'll ever have the ability to do this.

We do calculate DCC checksums for every message that hits our system. However, we have hesitated making that data available for mail filtering. Since our data stream can be easily polluted (e.g., install a honey pot, visit it yourself, sign the address you receive up for a newsletter) we're hesitant to use it for mail filtering without some human intervention.

If you look at the people with whom we share data (e.g., SURBL) this concern motivates a significant number of our decisions.

Re: Hony pots and checksum blockers

Author: B.White3 (26 Sep 07 4:24am)

Your concern seems to be over malicious people joining project homey pot, and then getting some news letters banned?

The way these DCC checkers work is that they count how many times they receive a report before they tell others that a given email is spam. A single report is not enough to get a good hit.

Are you sure your not being over cautios?