Thanks for the referral, that does look like a great solution, the problem is I committed myself to gmail about 4 years back, before which I ran my own mail server. I like the cloud convenience of gmail for not having to doggedly make backups, and the fact that I no longer have to religiously monitor my mail server uptime. Also I like their webmail interface.

It’s not a perfect solution (I don’t believe there is such a thing), but it works pretty well for me. There are downsides, all that desireable/undesireable content going into my inbox is one of them. I like that classification by the way, as it perfectly describes email content.

Whatever I do is going to have be something that fits in with my gmail paradigm. I’m not closed to the idea of migrating to my own mail server again, but the advantages would have to outweigh the admin overhead, which can be quite severe when you have mail server outages that need work immediately. DSPAM does go a long way though to being a solution that would encourage me to do so.

You’re such a luddite, shunning FB and Twitter for modern communication

A.

More seriously, though, I went through a similar thing something like 6-7 years ago or so. In my case, it wasn’t a case of technical limitations of email, but one of my mindset that nearly caused me to abandon email.

I then tried out DSPAM on my email account. It’s a purely Bayesian spam filter based on the same ideas as outlined in Paul Graham’s Plan for Spam.

The beauty of such a spam filter is that it doesn’t really classify your emails as spam/ham by some criteria imposed on you, but rather as undesirable/desirable by the criteria you provide to the filter based on your training.

It took me a bit to realize that I don’t really care about whether an email is ham, spam, or bacon & sausages with a side-order of spam, spam, spam, spam and spam. All I care about is whether it’s in my inbox or not. That makes such a Bayesian filter the perfect tool for me.

It doesn’t provide perfect accuracy, but when I checked just now it’s accuracy was at 98.859%. That’s 98.484% of undesirable emails quarantined, and 0.073% of desirable emails accidenally quarantined. I’m really very happy with those results, and can recommend DSPAM to anyone on that basis.

Also, it let’s me stay connected with people I like.