[NTLUG:Discuss] the breaking point of spam

[Kenneth Loafman]

Richard Geoffrion wrote:
> Something is wrong with the technology when 91% of an email server's mail is
> spam.  

I've been seeing the same thing for years.  One of the domains I admin
gets about 5,000 pieces of mail a day for a 7-man shop.  Of those, less
than 200 a day are real mail.  The rest are spam.

> Blacklists are NOT a solution. 

Some blacklists work, SpamCop for one.  Very few false positives.  I
reject about 90% of all mail using that and other lists (not SORBS).

> It appears that ComCrap is now
> incorrectly listing servers that forward mail for alias accounts as spam
> servers.  There seems to be no human contact at Comcrap but the automated
> utility seems to unlist multiple times upon request.

Interesting.  Are they trying out SPF to see if it will work?  One of
its major failures is in the situation you describe.

> Other issues where
> static/non-dialup IP addresses are listed in SORBS (a volunteer-run
> blacklist that should never be used by a business interested in receiving
> email) can leave end users unable to resolve issues on their own and leave
> them at the mercy of their ISPs to get off SORBS...which....ultimately
> leaves the ISP at the mercy of SORBS who is wholly unconcerned with prompt
> resolutions.

SORBS has always been overly aggressive and horribly slow to respond.

> I've recently taken the stance that it is the recipient's responsibility to
> receive email.  I have quit trying to fight the myriad of lists.

Unless you're emailing a newsletter or other bulk mail, they should not
be that much of a problem.  Bulk mail senders have some real problems.

> Where are people in the Linux community going to throw their support behind
> a replacement solution? AMTP looked good but doesn't look like it is going
> to be adopted.  Someone somewhere has GOT to think of something that can be
> implemented into the current setup that would make a cutover painless and
> desireable.

Bottom line is that if we don't watch it, we'll end up losing control of
the net to proprietary interests, i.e. Microsoft.  I seem to recall
seeing that they were working on a separate email transport not based on
SMTP.  All that needs to happen is for that to become a de facto
standard that has royalties and we're all screwed.

One real problem is that there are literally millions of mail servers
out there.  Cutover would take years and would have to be staged.

Another problem is anonymity.  If I want to send anonymous email now,
its possible.  A system that is capable of keeping out spammers would
likely not allow anonymity, and that's a necessity.

...Ken