--On Monday, October 09, 2006 16:28:24 -0600 Brent Kearney <[EMAIL PROTECTED]> wrote:

Some university campuses maintain strict control over their reverse DNS
entries, and so departments or on-campus organizations, research
institutes, etc. that run their own mail servers will have non-matching
forward and reverse DNS entries on their MX hosts.  Blocking rules like
this make life difficult for them as well.

I would submit to you that any university that doesn't reverse all internet-facing hosts needs to hire new people to handle DNS. It doesn't take a brain surgeon to write pointer records.

Having said that, read my response to Tim to see why that's not necessarily a problem. Or better yet, read the link I sent rather than assuming how the program works.

Cases such as these raise the question, if the blocked mail never gets
into your network, how would you know about the rate of false-positives?

Very simple. Anyone whose email bounces complains to me personally at a known good address on a separate domain. So far I've had one complaint, and I simply adjusted the scoring to overcome the stupidity of his ISP.

Obviously, my brief description wasn't meant to fully explain how policyd-weight works. Read the docs. It's far from a brute-force tool to reject email.

Paul Schmehl ([EMAIL PROTECTED])
Adjunct Information Security Officer
The University of Texas at Dallas
http://www.utdallas.edu/ir/security/

Attachment: p7s0OX3511CV0.p7s
Description: S/MIME cryptographic signature

Reply via email to