Kevin Werbach wrote a piece yesterday pondering the death of email. It’s not dead, but it’s gone from being a useful tool to something you have to cope with. These days everybody has email, everybody expects you to check and respond to your email, and at the same time, your mailbox is constantly filled with utter dreck by spammers, legitimate companies that you made the mistake of disclosing your email address to, “jokes,” urban legends, and email viruses.

So I’d say email is anything but dead. Everyone who works sitting at a desk is utterly dependent on it. These days, I use a two tiered approach to email filtering, and I read my email in two ways. On my hosting account, I have SpamAssassin installed. It’s not a perfect installation — for some reason it won’t process mailboxes for the Bayesian filter, it just hangs instead. But its rule based processing gets rid of nearly all my spam. I read my personal email using Mutt when I’m at work, and I see all the spam that gets by SA, which is pretty insignificant. At home, I’m running Thunderbird and using its built in Bayesian filter to weed out the spam that gets past SA. I have its filters trained well enough now that at home I don’t see any spam at all, and I have very few false positive problems as well, even for promotional emails that I actually want to get.

My work email account uses Exchange and Outlook, but I never have problems with it because I simply don’t use it for anything but sending mail to coworkers. I rarely use it to sign up for things on the Web, I don’t use it to send email to friends, and I’d certainly never use it to post to a public mailing list or Usenet. This insures that the email address never gets published on the Internet and thus puts me on the road to getting spam that doesn’t pass through a spam filter.

This is how I cope with email. The only serious improvement I’d like to make is fixing the Bayesian filtering in SpamAssassin, but to really get it working well, I’d need a key in Mutt that says “this is spam,” and I haven’t bothered to figure out how to set that up.