January, 11, 2003 archives

paul graham's latest on spam, will filters kill spam? just highlights the problem with how most people are applying filters. it would be better to reject email that you think is spam (preferably at the smtp level). this actually gets the spam out of your life, and gives a clue to the spammer that you are not willing to even see their crap.

what about false positives? i know that if i sent you mail, i would prefer that you simply reject it if you think it is spam. when you file it away in your spam folder, i have no way to know, and no way to correct the problem from my side. i just know i haven't heard back from you, and not whether it is because you hate me, or you just haven't looked in your spam folder, or you went through your spam folder and just overlooked my message. (or if services take paul's suggestion that the filtering be done automatically for the less clueful, that you even know that you have a thing called a spam folder that might have non-spam in it.)

i think you can basically draw a direct parallel between the pharmaceutical industry—focused on stifling symptoms on a case-by-case basis (for $20 a person, with new ways to squash the symptoms every year!)—and the spam-fighting industry.

on a far too regular basis, my mouse stops working. if i simply unplug it from the keyboard and plug it back in, it comes back to life. and there are frequent hiccups where i think the mouse has died, but it has not. it's practically a greek tragedy.

another random thing: sometimes after i've typed stuff into a text box in safari (like the one i'm using now) and then let it sit for a few seconds (or the focus moves?), the text inside the box reflows for no apparent reason whatsoever. very odd. it's a very peripheral-vision sort of moment, where you're not really sure that anything has really happened. (after playing round a bit more, it appears that the text is getting reflowed when i mouse over the buttons to submit the form. very bizarre.