I'm beginning to get the feeling that dealing with spam is a full-time job for most people. Despite creating dozens of filters and automating my mail management, the garbage mail continues to squeak through the cracks in my defenses, leaving offers for things I neither need nor want popping up in my various inboxes.

Part of the problem is the large number of email accounts I use. A contributing factor is that I use two different email programs--Outlook XP and Eudora Pro--to manage my email. Add to this mix the fact that a few of my addresses have existed for more than 10 years (one or two are plastered all over the Web in articles I've written), and you can imagine how many spam lists I'm on. This state of affairs is partly my fault: I searched on some of my older email addresses and found references to them on the Google UseNet archive ( http://www.google.com/grphp?hl=en&tab=wg&ie=UTF-8&oe=UTF-8&q= ) that go back to the early 1990s. If I can find those addresses so easily, you know that the spammers have long since harvested them.

Because I'm not willing to get rid of all my old accounts, I'm forced to find some other ways to defend myself from this onslaught of mercantile missives. As a small office/home office (SOHO) user, I don't run my own mail servers (although I've done so in the past), so the variety of server-based antispam solutions aren't suitable for me. My initial foray into the world of client-side spam control leaves me with the impression that three types of solutions exist: external email filtering and authentication services, a middleware layer that sits on a local machine between mail applications and ISP mail servers, embedded applications or add-ins for existing email programs.

As a public service, I plan to install and use examples of each of these client-side solutions for the next few weeks. I have representative software, but if you have a special favorite that you'd like me to look at, send me an email message and let me know about it. I'd like reader feedback about the criteria by which you think I should judge these solutions.

In my opinion, the two major criteria should be ease of use and effectiveness. These features overlap significantly; a product that's easy to use but that makes erroneous guesses about content and consistently blocks email that isn't spam isn't really easy to use because you have to double-check everything it does. Conversely, software that requires consistent user interaction and is rarely wrong is little better than filtering spam manually. A happy medium between ease of use and effectiveness is necessary.

I know that the 5000 to 7000 email messages I receive each week will be a challenge for the client-side products I've chosen. In addition to mailing lists that I subscribe to that generate hundreds of messages a week, I also get dozens of press releases every week from computer product vendors. The press releases fulfill many spam filters' criteria for spam, but they are legitimate mail for me. I'm looking forward to conducting what should be an interesting series of tests.