Spam
Since the beginning of this year, I have been tracking the amount of spam that SpamAssassin has been filtering. It’s sickening how much spam comes in daily. At the moment, it’s probably averaging around 300 spams/day. And that’s not counting how much my server rejects before it even gets to SpamAssassin. And it’s only mail for three people - myself and my parents.
The graph for October is lower, because between Oct 15 2003 and Nov 15 2003 (or about that long, not sure exactly), my mail server was down and wasn’t tracking the spam. But that also means that the graph for November is only for the last half of the month. I figure that it would have reached 6000 or more. At the rate it’s going now, I wouldn’t be surprised if 9000+ spams got filtered this month.
Thanks to a lower threshold than default (4.0 instead of 5.0), and Bayesian filtering (see the Wikipedia entry for more information), I haven’t gotten any false positives for some time now. And only the occasional false negative. But I just feed those to sa-learn and all is good.
Note that I have modified SpamAssassin slightly to do the actual tracking of spam. Right now, I haven’t really publicly released my small change, but I’ve given it to people that have asked. I’ll have to properly release it in the future. Along with the code generating the graphs on my page.
I really don’t know how people can manage to read non-filtered email.