On Sunday, Sep 22, 2002, at 15:18 US/Pacific, Phil Tanny wrote:
> White listing software would maintain a list of everybody we
> would like to receive mail from.   It would log in to an
> email server and download any emails that were from someone
> on our "white list".   It might then delete all the other
> messages.  Or better yet, it could send an autoreply to
> senders of all the rejected messages offering an apology and
> an alternative means of contact.
>
> The rejected senders could be directed to a web page form.
> This form could be coded so the destination address is not
> harvestable, and so that human interaction would be required
> to operate the form.

Ah - like TMDA (http://tmda.net/)?

If such a program becomes widespread enough to seriously hurt spammers 
I'd be surprised if it took a year for spamware vendors to figure ways 
to automatically respond to the current systems' autoreplies or web 
pages. The other problem with this is that the From address is 
forgeable - if I set my From address to something like 
[EMAIL PROTECTED] or a popular mailing list's signup address it's 
probably going to get through to most people. A little more thought 
suggests that trying (support|admin|root|MAILER-DAEMON) in the user's 
domain or reusing addresses scraped from the same domain will probably 
be even more successful (that autoreply offers a nice way of confirming 
receipt, so you can guess relatively quickly).

Cryptographic signing can solve the forging issue but it really needs 
the web-of-trust or aggressive use of x509 user certificates *and* 
revocation lists to be effective. The web of trust approach would 
require much wider deployment and better user interfaces than anything 
we have now; it also fails to handle widely separated users well. 
Besides, what fraction of mail would make it through if you simply 
restricted your mail to messages having a valid PGP signature of any 
sort? The client support just doesn't seem to be here at the level we'd 
need to make this usable on a wide scale.

There's one delaying tactic I'd expect to work (requiring the user to 
manually rekey a nonce shown as part of an image with a noisy 
background to validate a public key). Outside of the obvious challenges 
for people with colorblindness or bad eye-sight, I think even that 
would fail surprisingly quickly - if you can generate it, they'll 
probably figure out a way to fake it.

Still, it might buy us a few years to clean up rogue networks before 
Moore's law makes it easily reversible. I'm not sure operators of 
mailing lists or large sites would consider that a worthwhile trade for 
the hassle it'll create for them.

Chris

_______________________________________________
spamcon-general mailing list
[EMAIL PROTECTED]
http://mail.spamcon.org/mailman/listinfo/spamcon-general#subscribers
Subscribe, unsubscribe, etc: Use the URL above or send "help" in body
    of message to [EMAIL PROTECTED] 
Contact administrator: [EMAIL PROTECTED]

Reply via email to