Identicons vs. Autobots

I have discussed commenting systems before. A recent flurry of work is attempting to address some of the concerns. Don Park has invented identicons. I'll let you go read about them yourself, but essentially, an identicon is a visual cue to identity. It maps an IP address to an image. The idea is to help readers identify a person's comments in a group of comments, whether because they want to read those comments, or as a simple indicator that the comments may not belong to the person they think they belong to.

The title, aside from being a childhood reference, suggests that identicons could be a way of combatting comment spam. If the comment spam was not removed completely, perhaps its identicon could be learned as one to skip past when scanning the comments. Slashdot does something similar with their Friends and Foes system and I do find it useful when deciding what to read out of hundreds of comments. I'll for sure read the friends and the enemies of the friends to get the opposing view points.

The identicon idea does have drawbacks. For one, there are already a handful of me-too implementations that, while cool, contribute to the identity problem because you will end up with different identicons in different places.

In all, it's a very interesting solution to a very complex problem and is worth some thought.