Jim Choate wrote:
They do it now, sans encryption. The mass distribution is what makes it economical. If the encryption can be gateway'ed then it's useless and doesn't raise the cost significantly.
it is INDIVIDUAL encryption. want to spam 100,000 people? gotta encrypt your spam 100,000 times to 100,000 different keys. if you have a fast machine that takes only a second to do so, it'll cost you just over a day...
It's your problem, there are aspects of this proposal that are simply silly, and several others that haven't been adequately explained or examined.
You talk about decreasing the load due to spam, and don't even recognize that you've replaced it with a whole other process. One that potentialy could be more complicated, error prone, and expensive in time and resource impact than the original 'problem'.
this process is exactly as complicated as you sending me a PGP-encrypted message. remailer or not doesn't make a difference. don't you get it? all we do is couple remailers with PGP encryption, enforcing the second by a simple test in the first.
The solution to bad speech is more speech, not regulation. And don't kid yourself that setting up such a mechanism isn't regulatory.
it's my remailer, and I can do to it whatever I like. you are, of course, free to not use it.
What algorithm are you proposing to identify plain-text?
several have been discussed here over the past week or so. none are perfect, but most are good enough to distinguish between spam (which has a describeable structure) and encrypted stuff (which, too, has a describeable structure).
There are key managment issues, what is your proposal for this problem? There is the increased complication of admining the box (think of resources to support both the remailer operation as well as the encryption - consider that scale carefuly).
there is no additional complication. to prove that, I'm in the process of setting up just such a remailer. this takes some time since I have no experience in remailer operation, but it'll be done. if anyone wants to help me, send me a mail
You need the key to get into the remailer, otherwise how does it tell the message is encrypted?
I can see whether or not a text is greek or russian without understanding it. likewise, I bet you that anyone on this list can identify an ASCII-armoured PGP message with a single glance. shouldn't be too tough to teach that to a machine, should it?
You seriosly propose sticking some static PGP header for example will stop anyone, spammers know how to use word processors too you know.
please take a look at the simple script I posted here recently. tell me how you want to get it to accept any of the standard spam messages. hint: adding a fake PGP header won't work.