Monday, February 28, 2005

Spam, spam, spam spam

Spam is huge, inbox-choking, grandmother-annoying, time-strangling problem today, and it will continue to be so. A discussion of spam usually elicits everyone's favourite spam stopping software (whatever that may be). Free from having to view the apparently endless intentional misspellings of penis related pills, they praise their programs with great praise. , but as Andy Lester points out, filtering email based on content will never solve the spam problem. When I said the people who love their antispam software were free of it, I should have said they were freer of it than they were before.

As Andy rightly points out, the problem is not technological - aside from the inherent lack of accountability of most email - but it is philosophical. Imagine, if you will that you only receive ten emails per day. You have no time for spam, so you hire an assistant to screen out all the spam emails before you see them. As luck would have it, the economy has dictated your assistant is a temp - a new person each day. You never know how good they will be with computers, so you have to write out your rules for filtering spam, never knowing if the assistant is smart, or very very, dumb, like people who send chain emails or certain Yale graduates. There is no set of rules that cover all circumstances at a character-by-character level that will work. And computers are no smarter than certain Yale graduates.

The long-term solution is to have better email sending authentication (which by definition would have limited backward compatibility). In the short term, you must use the philosophically weak filters, stopping now and then to sift the wheat from the chaff. Remember if you spend one millisecond considering why you got any particular spam, you've already expended a vast amount of energy exceeding that which it took to create. But today, the problem is akin to trying to stop littering by hiring more garbage men.


Post a Comment

Links to this post:

Create a Link

<< Home