DOH!

mt-blacklist deleted a friends(tav) comment on here cause it contained the word F*u#C!K. I had the brain wave to quickly go to google and see if it was cached.. it was!! woo.. i’ll paste it in below.

o little spam! makes me jealous ;p

here are some thoughts on your good spam reduction techniques:

1. very good advice on the unsubscribing trick.

in addition, i would suggest being wary about clicking on urls sent in emails by friends. check them for parameters which might be referencing your email address, either through unique ids or sometimes even directly (foo.com?e=foo@blah.com)

i fell for this one in one of the “passion” greeting cards that a lover sent me. try urls without providing any of the extra variables – you can often see the same content without providing automatic quality control for spam databases.

2. i would say that most decent email harvesters used by spammers are now capable of handling most of the “human-readability” protections that you describe above.

here’s my preferred form of email obfuscation, which too would of course soon be assimilated into spam-tech, but for now is cool:

document.write(‘’)tav@espians.comdocument.write(‘’)

it makes the email address both clickable (even downgrades well to non-javascript supporting browsers!), copyable, and you don’t have to waste time explaining to the not so enlightened as to why you have NO SPAM in the middle of your email addresses…

3. this is a pretty cool trick used by spammers. simple, yet very elegant.

i just wish that more email clients had support for optionally turning off images. i have only seen ximian evolution do that.

4. if you are using qmail, and have your email account set up right, you can automatically make up as many email addresses as you please, with no extra admin hastle… e.g. tav-someshitwebsite@espians.com would work.

this way, i can:

anyways, as much as taking these measures is beneficial, it doesn’t solve the underlying problems with mail spam, which are that:

a) the way email works is crap.

b) there are some naive individuals out there, who obviously do fall for the spam, and thus do make it worthwhile for the spammers to waste the rest of our time.

c) this then affects us because, again, email is pretty poorly designed.

my thoughts on this are to have a complete reversal to how we approach messaging. instead of people sending mails to your address, you should simply check the outbox of your friends (those in your trust matrix).

does talon have a message for me? does john? does …?

of course, there is the question of how do you then get messages from people who don’t know you…?

there, a combination of transitive trust (i.e. does that person know someone that you trust, and be able to pass the message on through them?) combined with other methods (such as usage of reputation backed currencies) would prove pretty satisfactory.

and, failing that of course, one could revert to a system similar to the current system, but use some slightly better methods such as the bayesian filtering techniques that have been implemented as a result of this article on spam href=”http://www.paulgraham.com/spam.html”>http://www.paulgraham.com/spam.html

and, perhaps use a system of overlapping distributed seed data sets for the filtering?

starts off by checking against your personal profile, then the collective profile of those who you work closely with, then… all the way up to a global one.

anyways, need to run off now. hope it makes sense, if not, i’ll be happy to rephrase/elaborate.