Re: PLUG: PMAS

In article <5dpvr5F35i1raU1@xxxxxxxxxxxxxxxxxx>, bill@xxxxxxxxxxx (Bill Gunshannon) writes:
In article <TRHdi.110033$n_.88261@attbi_s21>,
"John E. Malmberg" <wb8tyw@xxxxxxxxxxx> writes:
Bill Gunshannon wrote:

Believe it or not, this is one of the upsides to my idea. It would,
hopefully, spawn some new businesses specifically to provide "clean"
email sites. But the reality today is that probably 90% of the current
MTA servers are run by clueless sys admins.

It is my understanding that all servers by RFC are required to have a
valid rDNS.

IP stacks and applications violate the RFC's every day. RFC does not
stand for law. Heck, you might say they're not even suggestions.
The name is "Request for Comments"!! :-)

Many of the RFCs are also STDs (ie STANDARDS) see

http://www.faqs.org/rfcs/std/std-index.html


David Webb
Security team leader
CCSS
Middlesex University


Several major e-mail providers have tried to enforce this check and
found that they could not because too many real mail servers that they
could not afford to block did not have the basic competency to set an rDNS.

I reject servers when there are conflicts between the IP address they
come in on, the rDNS mapping and who they say they are with HELO. Stops
lot's of SPAM. Has resulted in a few problems. I have told admins at
remote sites about their problem and had them fix it. But then, I have
also had some who refused to fix it. Their problem, not mine.


A strict rDNS check will eliminate zombies from being able to easily
pretend to be real mail servers, and will significantly impact the
ability to send spam.

Essentially the spam problem is from three points:

1. Network owners that do not act on complaints.

2. Network owners that do not configure their mail server identification
correctly according to RFCs.

3. Network owners that are willingly spam havens.

I would sum it up much simpler. Stupid managers and incompetent admins.
Of course, number two is usually the result of number one.



A new mail protocol will not eliminate these problems.

I am not proposing a new one. I am proposing an old one.



Some parts of this discussion are becoming repetitive. I have made
several points, that have not been refuted, mainly just ignored.

1. The state of the art of automated spam filtering is that over 99% of
spam can be removed with out impacting real messages.

I don't agree. I think much more real email is lost and the recipient
has no way of ever knowing he did not receive the email.


Just because there are popular spam filters and techniques that do not
do it right does not invalidate that point. It just demonstrates that
those filters are not state of the art.

I have seen nothing posted today or on any other day that disputes point
number 1.

When you don't receive an email, how do you know you didn't receive it?
That always reminded me of the places around here that offer rewards to
anonymous tipsters. Duh.......



2. There is a large misconception and a lot of posts on many forums
claiming high error rates of DNSbls. I have seen none backed up with a
real example from a mainstream used blocking list. In fact I have not
seen an example from an aggressive blocking list known for false

Unfortunately it appears here that perception remains with those that
previously held it, even though no one has produced any evidence to show
a specific real e-mail rejected for being in a mainstream used blocking
list.

Huh? I just told you yesterday that I had an email form some that was
rejected because a major site is RBLed. It wasn't SPAM it was real
email. I have a friend locally who send stuff to me at my Army email
address because he has a Comcast account and some number of their
servers are on various RBL's so he can never be sure his emails will
get thru. These are not SPAM, theae are real messages from major players.
How can you say that only SPAM gets stopped?



3. There seems to be a great concern about having a real e-mail
accidentally rejected because of a DNSBL, but no concern about it being
lost from other causes,

Other causes not related to SPAM have become highly unlikely in todays
networks. The net doesn't just loose emails. Servers keep trying for
several days before giving up.

including end user spam filters, which can not
notify the sender of misclassification.

But that is part of the same problem. Eliminate the need to filter so\
agressively and this problem goes away, too.

SMTP does not guarantee
delivery or notification of non-delivery.

That's true, but I owuld bett that the number of emails lost because of
having to deal with SPAM is orders of magnitude larger than the number
of emails lost through other methods.


The less effective spam filtering is, the higher the odds of e-mail
getting lost for other reasons. That is not being taken into account in
the criticisms of DNSbls use.

Your right. If it weren't for the hundreds of emails lost thru SPAM
filtering the five lost by other methods would look really bad. :-)

bill

--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill@xxxxxxxxxxxxxxx | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
.

Relevant Pages

  • Re: MTA on non-standard port
    ... one machine on my LAN is allowed to speak SMTP to the world. ... with his/her laptop, which happens to be infected and sends spam, etc. ... Comcast stirred up the ants. ... servers!), citing "an incident of spam from my IP address". ...
    (freebsd-questions)
  • Re: PLUG: PMAS
    ... MTA servers are run by clueless sys admins. ... ability to send spam. ... I think much more real email is lost and the recipient ... The less effective spam filtering is, the higher the odds of e-mail ...
    (comp.os.vms)
  • Re: MTA on non-standard port
    ... just to give you some idea: my home LAN has a FreeBSD box used ... with his/her laptop, which happens to be infected and sends spam, etc. ... Eventually they stated that I could send mail through their mail servers ... I've had two separate incidents of me sending mail to individuals, ...
    (freebsd-questions)
  • UPDATED SUMMARY: Simple anti-spam system using open-source software and freely-available data
    ... I run sendmail and have had excellent results using a layered approach ... to blocking spam. ... That responsibility rests with the people whose servers ... but the load I impose on the DNSBLs ...
    (SunManagers)
  • Re: ISPs blocking SMTP connections from dynamic IP address space
    ... >>ip blocking for legitimate servers is silly. ... Because spam ... AOL will likely cancel the ...
    (freebsd-questions)