Stamps vs SPAM

Posted 18 Sep 2002 at 21:16 UTC by Fare Share This

Having lots of much better things to do, I instead wrote up this article, Stamps vs SPAM on a plan to eliminate spam by requiring unidentified senders to pay a small amount in money or in time.

Though I am skeptic on the current feasibility of money payments, paying in time by passing a short and simple turing test at a website (or directly within your MUA, with the correct plugin) seems to me a promiseful way to screen off bulk mailers.

Actually, I'm not just trying to spread my memes with the article, I'm also trying to find people who'd be interested in implementing such email postage payment protocol and associated software. If you are interested, know someone who could be, or a place where to advertise my article and look for interested people, please tell me.

Hashcash, posted 19 Sep 2002 at 00:22 UTC by FarcePest » (Journeyer)

I'm surprised you missed hashcash. Hashcash is a burnt-offering of CPU cycles using partial MD5 collisions...

Paying More For Email?, posted 19 Sep 2002 at 03:30 UTC by Grit » (Journeyer)

I'm bothered by proposals of this sort because I don't really feel like I should have to pay more for email. U.S. postage keeps going up and up for the same service; in order for e-mail postage to be effective it must do so, too. Fundamentally, I want goods and services to become cheaper as time goes on, not more expensive.

Further, the fact that I someone has to pay to send me snail mail doesn't mean I don't get spam in the "real world". In fact, for me the ratio of good mail to spam is a lot lower for physical mail than my (6-year-old, SpamAssassin-filtered) primary email account.

Whitelisting is futile without some form of authentication; I've already been getting plenty of spam addressed from myself, or from mailing lists. I think this makes schemes like ChoiceMail ultimately futile. But, if we can solve the authentication problem (so that all email is easily trackable to a real person) then there's little need for postage because filtering or whitelisting can be effective.

Postage in Computer time vs Human time, posted 19 Sep 2002 at 11:37 UTC by Fare » (Journeyer)

Well, I did know about hashcash, and cited the CAMRAM project in a footnote. Since I've received quite some feedback about it, I moved it to the main text, and cited hashcash explicitly. Hashcash has many problems: first it won't stop robots, it will only slows them down (and users of good old machines will be more annoyed than spammers with brand new spamming farms). But most importantly, it requires the sender to install new software, and thus either cuts you from many correspondants, or else forces you to fallback to another system for all the rest of the mail. That is, you're still in trouble until everyone uses it, still in trouble afterwards (though allegedly less), and have no means to convince anyone to use the system. (Oh, and if payment is to be done by burning cycles, I'd much rather it be done for a good cause, like SETI@Home or some such, instead of aimless cryptographic collision detection.) I'm not saying hashcash is useless - but it does not suffice in itself to stop SPAM.

In contrast, paying in human time could be fully automated (given a proper RFC for robot responders), and would require no content filtering by the recipient. Recipients could experience inconvenience if they don't use whitelisting and authentication software - but the point is precisely that they should use such software. Senders would experience a little inconvenience, but that is precisely the point; and the fact each time they have to click on a website to confirm an email, they are reminded that they could avoid the inconvenience by upgrading their tools, while being saved from spam at the same time, is a clear incentive as well as viral marketing trick for them to adopt the system.

Yes, Paying MORE for Email, posted 19 Sep 2002 at 11:48 UTC by Fare » (Journeyer)

I think that monetary stamp payment would be quite good, and much better than payment in either human time or computer time, if only the problems with security were solved.

Most importantly, you'd only need advance money when initiating a conversation, and would be reimbursed in the first reply as a matter of politeness (with help from default settings of the postage software). Thus, the total credit you need to spend from your postage account would be determined by the number of unsollicited mail you send - which is exactly the point.

Of course, some impolite people won't pay postage back - and it's a great way to detect people to put in your blacklist. If your time is worth anything, you'll be glad you lost the postage instead of your time trying to discuss with a nogoodnik.

and why not a universal contract of internet use?, posted 27 Sep 2002 at 20:17 UTC by jul » (Master)

My opinion may be we should stick to a system of responsability of domain name owner. The RFC 1591 describes partly what is expected from them.

Maybe we could consider that each (sub)domain name owner has a contract with the internet community such as : Part I (spam) - whois database information should be valid. If not the domain name is considered as free. - no mail should be accepted from non fully qualified domain name.

Part II - we should avoid the spoiling of computer ressources (disk space, bandwith) when avoidable (no video sent to mailing list/UCE...) - adminitrators are not to be perfect, but it would be great if they were expected to try to find a solution with complainers (when there is not too much people). (obligation of trying) - if the administrator of the implied sub-domain does not answer to complaint, the owner of the domain just on top of it should have the responsability of trying to solve the problem, and maybe revoking the ownership of the sub-domain name. - a part of the contract that is signed by each internaut may be an Universal Chart of Computer Usage(UCCU). - we should reserve ourselves the right to deny routing and service access to domain that systematicaly enfringes the UCCU (death penalty enforced with RBL)

I guess companies that makes money from offering internet access may see their concern about spam raising if they were not only worrying about the cost of their (unused) bandwidth, but also of the cost of not being connected to internet. And some registrars might as well be more serious about their obligations if the ICANN was enforcing this "contract".

Maybe there is already an RFC on this topic :-)

universal contract? haha, posted 1 Oct 2002 at 16:21 UTC by Fare » (Journeyer)

you know, jul - you have been reading too much Rousseau. There has never been such a thing as a universal contract, and will never been anything like it, before the universe dies out. There are already ISPs cutting down on spammers. Spammers just create new accounts, or switch ISPs.

Computer ressource is cheap - human time is the ultimate scarce resource. The transaction costs involved in a contract are expensive - for a global contract, they are just inimaginably high. Not to talk about the costs to enforce such a contract, even if it exists.

Given suitable software infrastructure, such as being developed by e.g. the CAMRAM project, stamping has the advantage of being directly beneficial to its adopter even as a purely local measure.

Fare enough, posted 15 Oct 2002 at 13:29 UTC by jul » (Master)

Stamping was the price to put on a paper that would have to be delivered. Not only was it the expression of a full jacobinistic vision (one rules for all), but also the beginning of the control of information.

People are more at ease with a central authority that can bring «justice for all». But, if you accept stamping, you will accept publishing agreement for every website in order to struggle against pedophilia at first, politicaly uncorrect website then.

My point is it will not prevent spamming, it will just help the idea of a control of information over internet. For now on, I use spamassassin, and as spammer love to crawl webpages for emails, spammers are on a page of my website. This may be nothing, but I find it funny and better than nothing.

Je dis ça, c'est comme si je disais rien :-)

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page