Legislation, moral principles, or revenge won't solve the spam problem. But clever people with glasses might make it bearable.
It seems the spam menace is causing a lot more hand-wringing than it used to among the Internet public. A recent Harris Poll found that 80 percent of Internet users claimed to be "very annoyed" by spam. Ferris Research estimated spam costs US corporations US$8.9 billion a year, European corporations US$2.5 billion, and another US$500 million in costs to US and European ISPs. While I don't think you can ever get realistic figures by using shonky calculations like time taken to delete spam x number of spam messages x average wage, the message is clearââ,¬"there's a lot of it about, and we're starting to get pissed off.
But what to do?
There's a growing movement towards making spam illegal in the US. The Harris Poll found 74 percent of those surveyed were in favour of making bulk e-mail illegal. To me this smacks of US-centric short sightedness. While banning spam in the US might curtail the bulk mailing activities of legitimate US-based businesses, it won't stop the porn, the pyramid schemes, the Nigerian scams, and all the other crud, because these are not sent by legitimate businesses, and mostly not sent from the US. Although Australia can be pretty much guaranteed to follow Dubya's lead on anything, it's reasonable to assume there will be some countries that still allow spam.
Then we've got my mates the spam vigilantes, like ORDB and SpamCop, who think if they compile lists of known spam IP addresses or domain names, they'll be able to block a majority of spam. This doesn't work either, of course.
And then there's the option of installing software on your mail client or server, which in theory prevents spam from reaching your mailbox.
Over the last six months or so, I've been trialling a variety of spam filtering products. In a previous column I advocated the use of filtering above legislative or vigilante means, which did not sit well with a number of T&B readers. "You shouldn't have to pay for software to block spam!" they wailed. Which is fine as an abstract moral principle, but if my business is losing $8.9 billion a year, I'm not going to wait around for abstract moral principles to assert themselves.
I say the vigilante method doesn't work, because in my various trials, I've been able to compare its effectiveness with other methods. What I discoveredââ,¬"and I've been kicking myself for not understanding this earlierââ,¬"is that domain name or IP address filtering for spam e-mail is every bit as ineffective as domain/IP blocking for porn Web sites. Sure, it catches some spam, maybe 30 percent, but the majority slips past, and there's a high rate of false positivesââ,¬"messages flagged as spam that aren't.
That's because no list can ever hope to accurately and comprehensively cover all the spammers in the world. We know porn filtering by blacklist doesn't work; why should we believe it works for spam? Or are spam vigilante lists merely childish revenge mechanisms against companies the vigilantes don't like?
Thankfully, there is a method of spam filtering that works. Not perfectly, but better than anything else I've seen. There's an open source package called SpamAssassin, which uses a wide variety of heuristic tests to give each message a score. If the score is above a certain threshold, bang, it's spam. While many packages use this approach, the intelligence and subtlety of SpamAssassin's rules give it a real edge.
SpamAssassin works with a wide variety of Linux-based mail servers, while a company called Deersoft has created a commercial version that integrates with Exchange Server or the Outlook client. (Deersoft was recently bought out by Network Associates, and the technology will be integrated into McAfee SpamKiller.)
Although I'd rather my employer invested in server filtering software, the client software has drastically reduced the amount of spam that reaches my inbox. So while the rest of you are gnashing your teeth about moral principles, or sitting in the dark compiling lists of companies you don't like, I'm getting on with using my e-mail for what it's intended to doââ,¬"keeping in touch with my friends.
Subscribe now to Australian Technology & Business magazine.
I have used spamcop for over a year, and I find it very effective in cutting down the amount of spam I receive, basically a 90 percent reduction in the volume of spam. However you have to use it every day without fail or the spam will build up again. This combined with effective legislation will put a large dent in spam production. Dont ever expect to block it all, as long as you can get rid of 95 percent you are making great gains. And as for my hotmail account last time I checked 163 emails, 1 real 162 spam (but then who cares, I just treat hotmail as a dead letter address).