On Tue, Mar 04, 2003 at 08:26:05AM -0500, David F. Skoll wrote:
On Tue, 4 Mar 2003, wayne wrote:
Spammers don't care about undelivered spam. Most of the time, they
won't even know because they use open proxies, forged from addresses
and forged reply-to addresses.
However, there is a benefit to replying to spam with a 554 SMTP code.
If the spammers have been using an open relay, the open relay will generate
More to the point, while spammers indeed don't care about getting
bounces on underlivered spam, it is simply very poor software
engineering to design a system that simply drops mail on the floor
and informs nobody who would care about it. Any false positive in
such a system is a lost piece of mail with no diagnostic to tell
people about it.
Obviously the volume of spam makes it inappropriate to report errors
one email at a time the way you would for one to one mail. So the
challenge is to build error aggregation systems which allow people
to view digests or statistics on errors that are relevant to them.
(Ideally you also store the mails somewhere that they can be
fetched by the target if desired. Certainly disk is cheap enough
that if you have the message in ram, you can store it to a database
for a week or so for fetch-by-url. However, if your technique
depends on blocking the mail before it has even finished arriving,
this is difficult, though you can still store as much as you did
receive. Simply appending a block of data to disk is not so expensive
an operation.)
Reporting errors back to legit bulk mailers -- again, in aggregate --
is easy because they use legit headers and a unified from. In
fact most MTAs do this already.
Giving up the principles of good software engineering and protocol
design, so that messages don't vanish into black holes with no
diagnostics should be one of our last-resort choices. There are many
stories from the anti-spam wars of people whose (valid) mail was just been
silently discarded who only found out about it by accident.
And let's not forget that open relay operators are victims of the
spammers. It may be necessary to close those relays, but they are
still the victim, not the perp.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg