On Wednesday, January 21, 2004, at 01:12 PM, James M Galvin wrote:
Just to be clear, when you use "anonymous" you truly mean anonymous in
the sense of the "remailers" that we used to hear so much about.
Would you permit "site traceability" as compared to "author
traceability" or is that too in conflict with anonymity?
Well, I'm cynical enough to believe that both perfect anonymity AND
perfect tracability are impossible. So we're always talking about
degrees of difficulty. In practical terms, I think the ideal would be
a level of anonymity in which a message could easily be traced to the
first "reputable" ISP that permitted the message onto the net, but the
degree of tracability from there would be highly ISP-dependent. An ISP
(or anonymous relay) that wanted to be known as unfriendly to spammers
but friendly to anonymous email would have to implement some special
measures to keep spammers off *their* system, but I'm not convinced
that it would be impossible, by any means -- and much easier for a
single service than a standards body. So, yes, imagine a traditional
anonymous remailer augmented with additional, possibly proprietary
mechanisms to deter spam.
My point is that there are certain types of message that should be easy
to trace to their senders, and certain types that should be hard to so
trace, and we should do our best to accommodate both. The relevant
variables include not only the originator but also the volume of
messages from the same source, and the preferences of the recipient.
By the way, I've just become aware of a clever new idea for an economic
solution to the spam problem, and it's not just another recycling of
the idea of charging postage for email. It's called "Attention Bonds"
and is worth a read for the long-term thoughts it provokes:
Cheers. -- Nathaniel