The Spam Wars.

Wendy Grossman's "The Spam Wars" (November) was interesting and informative. However, I believe that the economic methods for fighting spare were given short shrift. One of the reasons the "penny per e-mail" method is too often dismissed is that there's not much discussion of exactly where that penny goes. The sender pays a penny, but who gets it? The right answer to that question can mitigate some of the problems Grossman mentions: The penny should go to the owner of the ISP that delivers the e-mail to the final recipient, so that, in effect, an ISP charges others to deliver e-mail to its customers.

Grossman objects that no ISP is set up to charge this way. But 10 years ago, no ISP in the world was set up to validate relayed e-mail or to filter spare. Now both practices are common. This required changes to mail-server software, and adding a "sender pays" system would not be much different.

She argues that it would require an entirely new infrastructure for the industry. But we already have most of the pieces of such an infrastructure, including Internet credit card payments, Internet bill payments, PayPal, and similar services. I've just noticed that even Yahoo! will let me attach money to this e-mail via their "Yahoo! PayDirect from HSBC."

Grossman also worries that "sender pays" would kill free services such as Yahoo! and Hotmail. This is simply one additional cost added to their existing costs. Yahoo! and Hotmail already have to pay for computers, hard disks, bandwidth, and personnel. It's all funded from their advertising. However, at the same time, it is an additional stream of income. Yahoo! and Hotmail will receive a penny for each e-mail they deliver to their customers. They'd probably throw away all their spam filtering software the day such a system is put into place!

Further, the penny doesn't have to be paid immediately upon receiving e-mail. The larger ISPS could easily set up accounts where payments could be made monthly based on the actual auditable number of e-mails. This easily solves the issues of "fractional penny" payments and addresses concerns about massive additional e-mail traffic for processing payments.

Would such a system make mailing lists economically unfeasible? Any mailing list can simply require that each recipient send an e-mail back to the main server for each e-mail they receive. The list can send out 10,000 e-mails to 10,000 customers at a cost of $100. Each of 10,000 recipients sends an acknowledgement e-mail back, so that the list then receives 10,000 pennies.

Any recipient who fails to perform this duty, which they'd have to agree to as a condition of joining, is dropped from the list. It wouldn't be long before mail client software would have a "penny payback" system in place, where the user can control exactly which recipient receives acknowledgements automatically.

A final difficulty Grossman raises is that the system would have to be legally mandated, but can't be mandated worldwide. But this does not need to be legally mandated, although legislation would certainly speed its acceptance worldwide. The Internet community already has numerous standards bodies that create rules and protocols for all sorts of interactions, such as the correct way for two mail servers to interact in sending each other e-mail. The millions of mail servers out there are perfectly free to follow those rules or not.

The ones who break the rules too severely end up being avoided in one manner or another. The "black hole" lists for open relays, for example, already have a large number of ISPS refusing e-mail from anyone who fails to follow the rules concerning open relays. No legislation was required for this.

Market forces would bring most of the Internet community on board eventually, although U.S. legislation would surely give it a jump-start. In the meantime, the customers of ISPS that use the method can have an address book of those for whom they'll "spot" the penny. Senders not on that list can go into the "no penny" folder for later review. This is better than the black hole system, where senders unfortunate enough to share an IP address with a known spammer cannot send e-mail to half the Internet! Instead of e-mail falling into a black hole, the sender would receive a polite note requesting $0.01 to deliver the e-mail.

All in all, I think the method is worth additional investigation. The only people who really can't afford a penny per e-mail are the ones who send out millions per day: spammers.

Alvin Sylvain

Temecula, CA

I know Wendy Grossman has put tremendous thought into the community-based solution to the spam problem, but I have to ask one question: Why is the burden of controlling unsavory activity to be placed on the community? In the offline world, isn't that what we have laws for?

I am one of those libertarians turned rabid anti-spam law advocates precisely because letting spammers continue unfettered is anti-libertarian. "Do what you want unless it harms someone else" is a libertarian credo, but spam clearly harms other people. Even if a technology-based solution works flawlessly, I still must advocate laws and enforcement, simply because spam harms others.

Spammers steal other people's resources: bandwidth, server capacity, whois databases, netadmin time, time spent deleting. When people are stealing, we don't solve it through benign vigilantism. We call the cops.

Mark Metz

Hollywood, FL