Today’s Internet is a junk-mailer’s wet dream. You can mail out as much as you want, and it’s all free! Twenty million people, all of whom have some sort of access to a computer, and you can reach every single one of them. There’s only one catch: they’re just as free as you are to mail whatever they want. If one thousand people mail back saying how much they’d love to buy your product, and 19,999,000 people mail back saying that they don’t want your junk mail, you’re out of luck. You just crashed every computer your Internet provider has. If you’re really lucky, they won’t sue your sorry hide.
Junk mail on the Internet is a form of spam. The origin of the term is lost in pre-processed history, but the meaning is any mailing that goes to a lot of people with no apparent purpose. In other words, spammers don’t check to see if the recipients are likely to want the ‘spam’. Spam is a little controversy on the net right now, but if the problem isn’t solved, it will become a big controversy later.
Junk mail on the Internet is a bit different than junk mail in real life. On the Internet, people pay for the mail they receive (if they pay at all). Nobody pays for any mail that they send. Imagine if Carol Wright didn’t have to pay for any of their mailings! They’d send something new out every other day. That’s what could easily happen on the net.
The simple solution would be to start charging for sent mail instead of received mail. But the Internet is becoming big, and nothing big ever went for the simple solution. Also, it’s unclear that this would even be possible. The post office can charge for mail and know that it’s been paid for--there’s only one post office. There are all sorts of Internet providers, and they would all have to agree, at least with their ‘neighbors’ on the infobahn, to whatever payment scheme is devised.
There are all sorts of strange ideas being floated around, from making spam illegal, to setting concerned individuals behind ‘spam firewalls’ that don’t let any spam inside. Here’s a sample “acceptable use policy” that came across during our discussion on the Internet Engineers Task Force mailing list:
The ISP reserves the right to remove access to any user or account at any time for any reason. All decisions by the ISP regarding account removal are final and not subject to appeal (Insert refunds detail for that month’s usage here) (?)
Remember what we’re talking about here: something as necessary for everyday living as the U.S. Postal Service or the telephone. How would you feel if your local baby bell suddenly, with no warning, removed your telephone service, and refused to talk to you about re-instatement? If you’re an individual, you have lost all contact with your friends and professional acquaintances. If you are a small business owner, you are out on the street. You can’t do business without the telephone. And when the information highway is as universal as the telephone, the above “policy” will be about as popular as a space heater at the snowman’s lodge.
You have to remember, though, that these are computer geeks. They’ve never been popular, and they don’t care for the idea now. The popularity of the net confuses the hell out of them. And they want it to confuse you:
Let n be the number of separate copies of a message sent within some time period. For k from 1 to n, let gk be the number of newsgroups to which each copy is cross-posted. (Treat each copy to a mailing list as having gk=1, because cross-posting between mailing lists doesn’t work.) Define a cost factor C, with C=1 for a single message sent to a single newsgroup or mailing list, and with higher values of C for multiple messages and for cross-posted messages. Something like this might work:
n (alpha-1) .--- beta C = n > g ‘--- k k=1 If all the gk are identical, that reduces to C = n^alpha*g^beta. (?)
Simple! A child could figure it out. Now all I need is a child to explain it to me. Forget Internet Service. This genius has a rewarding future in the Internal Revenue Service.
Spam firewalls are probably the best way of dealing with unwanted mail. It puts control over your mailbox in your hands. Not all Internet mail services charge per mail received. If I were to join a service that did charge per mail received, I’d make damn sure they gave me the ability to choose which mail I really wanted and which mail I didn’t want, before they charged me. But then, I don’t accept CODs unless I know what’s inside them either. It’s part of being a cranky old man.
America On-Line realized they were faced with this problem soon. Their members were beginning to complain about spam, and it was only a matter of time before they started to demand the ability to reject spam. That’s the kind of thing that has commercial services waking up screaming. They want to send you spam. With all the competition from Internet-only service providers, the only way they can make money is by selling your soul to an advertiser. But if you can--as any Internet-only user already has the ability to--filter out spam, you’ll never see any of AOL’s spam either. So AOL took the offensive and unilaterally started blocking “spam havens”. If they decide that too much spam is coming from an Internet provider, they block all mail from that Internet provider. So if one person on your provider sends out spam, and AOL catches them, you can no longer e-mail your friends at AOL.
Add to that the fact that their blocking software is just as buggy as the rest of AOL software, and you’ve got a major problem in the works. My current “main” Internet Service Provider is CTS. A few weeks ago, some CTS members noticed that their mail wasn’t getting through to their AOL friends. A little testing indicated that if the spam blocker was turned off, mail went through. If it was on, mail didn’t go through. So they looked on the list of blocked sites... and CTS was not there. A CTS technical wizard called AOL, who verified that CTS was not on the list of blocked sites, and if AOL was still refusing their mail it was CTS’ problem, not AOL’s. It took a day of testing by our service provider to prove to AOL’s technicians that the problem was with them. And then they apologized for having put CTS on a list that CTS had never been on in the first place, and things began to work again.
Most likely, the solution to spam will be a combination of all proposals. The biggest “solution” will be from the spammer side--they’re going to have to stop making people angry. Otherwise they won’t be able to sort the customers from the chaff of angry incoming mail. Some spammers are now refusing any replies. They’re requesting that customers call them on the telephone. This will only work for as long as the telephone is still a valid means of making purchases, and even then it loses those folks who hit the “reply” key automatically.
The second biggie will be from the recipient’s side. We’ll get to ‘filter’ our incoming mail. As networks themselves take over functions from computers, we will be able to tell the network to not even bother giving us mail from sales.com. Of course, if we embargo all mail from advertisers, we won’t know when that television we need goes on sale. And despite all of our complaining, we really do want to receive junk mail. Some of us do, at least, because some of us actually place orders from advertisements we received as junk mail. In the real world, junk mail pays.
When things sort out, we’ll have a compromise. It’s almost certain that the consumer calls the shots on the net. Explaining their decision not to provide traditional “pop-up” advertising on their service, the president of consumer resources for America On-Line said, “I think the presumption that a vendor will make the rules in a market where the technology empowers the consumer is wrong.” (?)
Some advertisers will, of course, be incorrigible spammers, sending cheap advertisements to every nook and cranny of the net. And we’ll tell our computers to ignore those advertisers. Other advertisers will try to make their advertising a little more appropriate, a little more informational, and a little less hard sell. And we’ll tell our computers to let those advertisements through, unless it’s about a topic we have absolutely no interest in. (Others, more ornery like myself, will tell our computers to stop all advertisements regardless, unless it’s about a topic we currently have an interest in.)
Imagine telling the post office that if they don’t stop dropping off all of those unsolicited advertisements, you’ll switch to a different postal carrier. That’s your option on the net. If you don’t like the service you’re getting, there are a dozen other providers who want your business.
This freedom brings with it problems for those who don’t like having to take responsibility for themselves. It means, for example, that there really isn’t any way to make it impossible for advertisers to send out advertisements. Any Internet provider who makes a policy against spam ads will see their ‘spammers’ simply move to a different provider. If the entire country makes it illegal to spam (and ‘spam’ is going to be a hard thing to define), spammers will simply move out of the country, taking their business and tax dollars out of whatever country was crazy enough to legislate against electronic mail.
This possibility of freedom is a problem for service providers who want to control their users. Places like America On-Line and Compuserve are used to having control over what their members see and don’t see. If they want to send an advertisement to all of their members whose interests include “breasts”, they don’t want their users automatically refusing those messages.
In September, 1996, America On-Line began its spam blocking in response to a spam organization called “Cyber Promotions Inc”. This is as if the post office were to suddenly stop delivering any mail from K-Mart. AOL’s users had no direct say over whether they, individually, would like to receive those promotions. Even now, they can only choose to live with all unapproved spam, or not receive any unapproved spam. It’s an across-the-board block on all messages from that provider. (And in general it doesn’t work, because spammers find it easy to change their address.)
I have little doubt that AOL did this because they knew their users were complaining about spam. I also have little doubt that they were desperate to come up with a way for their users to be able to block Cyber Promotions without being able to block America On-Line.
Under Unix, if a person were to decide they no longer wanted to receive any e-mail from “cyberpromo.com”, they would add the following line to their “filter” file:
if (from = ‘cyberpromo.com’) then deleteNot particularly easy to use, but a lot easier than being taken to court by Cyber Promotions. Unfortunately, you can just as easily add
if (from = ‘steve_case@aol.com’) then deleteto the same file. Steve Case, the owner of AOL, probably sees this as a flaw.
The biggest problem with ‘supply-side’ attempts at stopping spam are that they’re sure to backfire. The spammers are business accounts. They’re where the provider is making money, or will be, once businesses start to advertise on the net. Even if your Internet mailbox is with the same provider that the spammer came from, you don’t have as much clout as they do. And if the spammer is a business account with a provider other than your provider, your clout has been reduced to nil. I can’t see any way of dealing with spam that doesn’t involve individuals taking responsibility for what they want to receive and what they don’t want to receive.
As long as the government doesn’t get involved, the problem will take care of itself. Once the government gets involved, they’ll regulate the Internet so that it works at least as well as the post office. And I don’t think we want that. In fact, if we grant control over our own mailboxes to anyone else, we’ve already lost. There is money to be made in advertising, and any policy or legislation that ‘stops’ advertising will be repealed in favor of money. And everyone who has been counting on outside forces to keep their mailboxes empty will be out of luck. Those same outside forces will now keep their mailboxes full.
The short but fiery discussion of spam on the Internet Engineering Task Force mailing list was brightened by the following post, which may have been a joke, or may have been serious:
It appears that we have reached consensus that each list have a clear AUP which is FTP and WEB-able, a fixed fee per recipient for violations, and a note such as “violations of AUP costs $$$ per recipient” in any compilation of mailing lists.I originally thought Bill was being serious, but he later defends his claim of “consensus” by pointing out that five other people have said the same thing. “Or do you think that every reader of the list should vote?” Technically, of course, if you want a consensus you do want every member of the list to vote, or at least every member of the ‘spam’ discussion.The fee would be collected from the provider for the mailing list, who in turn would collect from the provider of the user who violates the AUP. Subscriber-Provider and Provider inter-connection already involves a contract of some sort, so this should not be a problem.
Suggested values for $$$ are currently $50 and $100. The local and remote providers retain 25% each for collection.
Now, somebody write this up. It should be taken up by the User Services Area. That’s what RFCs are for. I’d volunteer, but....
Bill.Simpson@um.cc.umich.edu
But it doesn’t matter, because this is the Internet. If those five people can describe their plan and come up with software and a consensus across the rest of the net, they don’t need the consensus of the IETF. And those five people include Eric Thomas, author of one of the major mailing list server packages on the Internet, and Ed Krol, one of the “granddaddy” Internet writers. Maybe they will be able to make their solution stick. Given the tone of his letter, it is unlikely that Bill’s solution will get very far. Yelling “Now, somebody else write this up” over the Internet is about as likely to work as doing so in the real world.
I think their chances are pretty low. The costs they’re suggesting are beyond anything one would suggest for advertising in the normal media. And the very existence of their plan will provide incentive for ‘undercutting’ it. The incentive will be at the provider level as well as the moderator level. There is neither a definition of “spam” nor a definition of a “mailing list”. In fact, there isn’t even a definition of a “moderator”, which will be crucial to the plan. Some mailing lists have many moderators; others have none. Spam is “un-asked for advertising”, but unasked by whom? Most spam is spam alone, but it’s possible to have spam and steak: advertising ‘hidden’ inside an otherwise informative message. I know someone who puts an advertisement for their comic book at the bottom of every message they send out to the net. Is it spam?
But in the end, it comes down simply to the fact that there is money in advertising, and any plan that attempts to block advertising completely will be beaten into the ground by that money. That, too, is the way of the Internet.
There are other solutions, some of them, I think, much more sound. For example, one of the mailing lists whose spamming brought about this discussion is a moderated mailing list. That means (in this case) that every message sent to the mailing list goes first to the moderator, who must approve of the message before it gets sent out to the rest of the membership. The spammer, however, supposedly faked approval, by fooling the mailing list software into thinking their advertisement was, in fact, approved for sending. This is more easily dealt with: if the same person were to get into the University of San Diego by pretending, to the security equipment, to be me, it’s called breaking and entering. Extending the same semantics to the net requires no special software or novel concept. In other words, as long as people take reasonable precautions against being ‘spammed’, attempts to break through those precautions are as obviously wrong as attempts to break through the lock on your front door are obviously wrong.
I’ve already mentioned firewalls. Internet service providers could easily provide ‘filtering’ mechanisms, wherein their customers can say things like “don’t ever give me mail from so-and-so”, or “always let my girlfriend’s mail through”. These ‘filters’ can even filter out bad mail based on the content of the messages. For example, I might say “don’t send me anything that talks about underage sex,” or “don’t send me anything from sales.com unless it’s talking about pedophilia.”
What likely isn’t going to be an option is the equivalent of postage stamps. In other words, charging for sending mail. There is talk about doing this, and very likely some form of charge will be implemented. But just like ‘real’ mail, this charge will be divided up between the ‘classes’ of mail: first class, special delivery, and bulk. And the costs of ‘bulk’ mail will be cheaper than ‘normal’ mail, thus making spams less costly than sending a message to your mother. Some versions of this charging scheme even include having little ‘agents’ tacked onto the mail message. Whenever the message has to cross over into another Internet provider in order to reach its destination, the ‘agent’ will negotiate with the new provider for a good rate. If more than one provider will get the message where it’s going, the agent will pit one against the other for better rates, trading off things like speed of delivery for cost, or vice versa. Under this scenario, the Internet becomes a vast robotic medieval bazaar, complete with haggling and shouting.
The need for good advertising space is definitely on the rise. Over the last month, I’ve seen advertisements for Russian women get rejected from two of the mailing lists I sponsor, and I’ve been asked by numerous businesses to “link” from my pages to theirs.
I was told I should advertise ski reports:
Date: Tue, 11 Jul 95 16:04:07 PDTI was told I should advertise magazines:
From: admin@aminews.com
Subject: Outdoor NewsJerry,
Link to *FREE* Non-Commercial Ski and Recreation News from AMI News, the nations largest producer of recreation and outdoor news and information.
This is the same news that is fed to ABC Radio Network,NBC Radio Network, and over 400 radio, television, and print outlets.
Over 150 Colleges and Universities across the country currently link to this information.
Date: Thu, 6 Jul 1995 13:18:09 -0400And I was told I should advertise titty bars:
From: reedo@hearstnewmedia.comFrom the people who bring you Popular Mechanics, Esquire, and Marie Claire... We would like to request that our recently-launched Newsstand of the Future be offered as a hyperlink in your service. We already offer a daily cartoon strip -- Secret Agent X9 -- that we feel fits in with the theme and image of your popular web site, along with frequently updated lotto information and tips, trivia contests, and inexpensive magazine and home video sales.
Web surfers who come to our digital kiosk can also win 1000 free lottery tickets, Harper’s Bazaar T-shirts, and lots of other fun and funky prizes. Over the next few months, our service will evolve and grow into a quirky, offbeat, and way-cool area full of vivid color and nifty features! We believe a relationship between our two sites would help both of us and we look forward to working with you on-line.