Image via WikipediaOver 90% of the email I receive is spam. Some days it’s as much as 100%. A few years ago that would have been a shocking statement, but in the 21st century it’s just a fact of life.
I’ve been a long term user of Spam Bully, but I recently uninstalled it. Axaware got too eager monetizing their product – I’m not interested in renting your software from you while you sit back and watch the money roll in.
I also report all spam I receive using SpamCop. Their UI is unfriendly to non-technical users (hell, it’s just unfriendly) but I enjoy having the opportunity to have spammers’ ISPs switch their Internet access off. Of course nowadays the computers actually sending the spam are probably just zombies, but even switching them off is progress.
The company I work for uses GFI MailEssentials, with mixed success. I can’t for the life of me get it to learn when it makes mistakes.
A while ago I moved my personal email over from POP via SpamCop to Google Apps. The GMail spam filter is very good but not perfect.
My contention is that spam blockers are completely unnecessary. SMTP was created with the assumption that you knew the people you were conversing with – it doesn’t account for spam at all. This basic technology needs to be rebuilt using assumptions that are true today.
Various attempts have been made to eliminate spam by bolting on new ideas like SPF and Sender ID, but unless everyone starts using them (which they haven’t), they won’t work (which they don’t).
The problem is that everyone’s trying to avoid touching SMTP – every mail program uses it, and it has worked reliably for decades. I think they need to just bite the bullet and get rid of SMTP, replacing it with a v2 that from the ground up accounts for spam. You’d have one generation of mail programs that supported both standards and then you could switch over permanently.
The prospect of a spam-free inbox would convince all users to upgrade without complaint. Also a huge number of people are using gmail/hotmail/yahoo accounts, if those companies moved over to the new standard and it would take care of a large proportion of users.
The new standard needs to be non-partisan – any whiff of license fees or patents and you’ll get less than universal acceptance, which will be useless. This probably means that no one company can develop this, which means lots of groups and companies will need to agree, which will stretch out the process. The thing is, to my knowledge they haven’t even started…
