Oh, what a tangled web we weave

I am not a normal person.

I am what the technology demographers refer to as "an early adopter." I like to try things out, to play with them, especially if they're new and shiny. In the past couple of years, social networks like MySpace, Facebook, and a horde of others have become all the rage on the Internet. There's no surprise there, actually; what these networks do is to simply mimic and extend the existing (non-electronic) social networks that we, as humans, participate in. Instead of meeting someone for coffee, we can exchange updates online. Like most activities related to using computers, there's nothing new here, it's just that things are made more efficient.

Last year, people started building aggregators on top of the existing social networks. FriendFeed is the classic example here: it creates a new social network for you by aggregating your content posted in dozens of other social sites across the web. In an earlier blog post, I wrote about the two types of social networking sites: services and connectors.

YouTube provides a service: it lets me share my videos. Flickr lets me share my photos. I don't want to have to share my videos on Flickr, and YouTube won't let me share still photos. But, on FriendFeed, I can aggregate content from both YouTube and Flickr into a newer, wider social network.

However, there's a new brand of connector that has recently come over the horizon, and which I believe is a notable development. In addition to merely aggregating services, FriendFeed now lets you (in at least a limited manner) generate content for the other services. Specifically, if you post an item on FriendFeed, it will repost it on Twitter.

This meta-service is not without its problems; there was an incident last week where a FriendFeed user's comments got posted to his blog via Disqus, which then posted the comment back to FriendFeed, which Disqus then posted back to his blog, which posted to FriendFeed, which created an infinite loop which only terminated when he realized what was going on and that he had more than 1,000 identical comments on one FriendFeed posting.

More complex problems can arise; for example, it's theoretically possible to have Twitter update your Facebook status, which then gets fed into FriendFeed, which automatically posts it to Twitter, which updates your Facebook status. (I've tried this one; luckily, one of the services along the way was smart enough to detect duplicates and stop it before the universe imploded.)

While some of the concerns with meta-services are technological (such as detecting duplicate), others involve more human factors: where do I want my status to live? Is it for the world to see, or only for a select group of friends? Do I want my coworkers to see my LOLcat photos? Do I want my high-school buddies to hear about my work issues? Right now, there's no control over the content in this space, so it's up to each user to individually control who sees what.

This is a perfect situation for a standards-based mechanism for communicating social network metadata. Open Social is attempting to do something like that, but they are fundamentally misdirected and will fail. They cannot expect closed, proprietary services like MySpace and Facebook to give up control of their most valuable asset, their social graph. Instead, what's needed is a simple mechanism for transmitting social controls along with the content. Since the vast amount of social content is available via RSS, I'm assuming that there will be a namespaced extension to RSS to permit these types of disclosures.

Unfortunately, until there is a major embarrassment caused by the failure of one of the social networks to keep something private, there will be no movement on this in the near future.