On Wed, 2003-03-05 at 13:01, Richard wrote:
Thanks for taking the suggestion seriously enough to respond. However:
1. Yes but the ISPs are detecting millions of spam at the moment and there
idea doesn't rely on them detecting every single one.
No, but the possibility of being wrong about just one creates problems
and costs far beyond the act of just passively accepting the spam. We
are talking about actions a company will be taking, not some Joe off the
street with a dialup.
2. Once and they are being told that their emails are being seen as spam
which is a good thing.
Who being told what? You are assuming that an email in a spam actually
maps to a real person somewhere? You haven't been looking at enough
spam.
3. Why? they are using up no more bandwidth than if they were just
forwarding the spam unfiltered. If the idea worked it will cut the amount of
spam and therefore free bandwidth.
This is ADDITIONAL bandwidth on top of that of the spam coming in.
Sending an email is a single event, an attack has to last for much
longer to be of any "use".
4. This is no more vigilantism than is the unilateral filtering of the
spam in the first place.
I completely disagree. Per user filtering is in no way like using a DOS
attack against an IP you believe belongs to a spammer. Unilateral
filtering isn't either. Filtering is not an attack.
5. How?
There is some criteria for determining what is the spam that you have to
respond to. All someone has to do is know what that is and then use that
to generate an email that causes you to "attack".
Peter
Pinging every spam link discovered would dramatically increase the
bandwidth the spammer was using and therefore their costs.
Problems:
1. How to "detect" the spam in the first place. Didn't you just say
above that spammers will continue to outwit filters? Detection is
essentially the same as filtering in this case.
2. Given a false positive, we would be pinging or emailing a valid non
spammer IP. This is called a Denial Of Service. People get sued for crap
like this.
3. For a large ISP, the outbound pings, emails, whatever to an IP deemed
to be a spammer would quickly cause a problem on outbound links.
Bandwidth costs money.
4. Vigilantism will create more problems than it solves.
5. Could easily be used for a distributed DOS against a person or site.
6. There was someone who suggested this at the recent spam conference in
Boston. It wasn't a popular suggestion for the reasons above and more.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg