Top

Spamnix for Eudora evolving

Eudora has always been my favorite mail application. It (almost) supports standard mailbox format (one mailbox = one file) so its easy to port mail from one platform to another, if configured right its virtually bulletproof against viruses, and it just works. My one gripe with it is that its only available for Windows and Mac.

I have written about how Eudora 6 will make it even easier to battle spam with Eudora, and I have to admit that when I saw in which direction Qualcomm is going with Eudora, I was also wondering how it would affect companies such as Spamnix that supplies anti-spam tools for Eudora.

While Eudora 6 is being designed to allow easy implementation of third party anti-spam applications and services, it is also taking a slice out of the market with a very simple anti-spam tool included in the mail application itself. One that should work for most people right out of the box. The little tests I’ve done shows that the anti-spam feature of Eudora 6.0 has a hit-rate of over 90% straight out the box – without any work needed. Pretty impressive if you ask me, and the added benefit of a tool that is hooked right into the mail application, makes it easy to white list everyone in your address book.

So how does Spamnix meet the future competition, by giving up? Not at all. Today, in an early beta/alpha release their new version sports more muscles than ever with the inclusion of a Bayesian

machine-learning system. (And I was just wondering when they would come around to that!). According to Spamnix software they have archived a 99% accurate spam filtering with this version already. That my friends are almost as good as a challenge-response system such as the one implemented by Earthlink earlier this week, but without all the disadvantages (with senders having to verify themselves before their mail goes through).

Look out for public betas of Spamnix for Eudora 1.2, coming for Windows and Mac OS X soon!

Psst: David Mertz has written a good article about spam filtering techniques, among them Bayesian filtering – if you are wondering what it is.

Comments are closed.