Just another predictive stake in the ground, this time regarding Twitter. Unless it federates it will fail.

While Facebook has recently announced 200 million users world-wide, and Twitter, who does not release official user population figures has maybe eight million, the chances of the latter reaching or surpassing the former are slim.

Why can't Twitter do it if Facebook already has?

Because Twitter is more like SMTP, the simple mail transfer protocol, while Facebook is more like AOL. Or to put it another way, you could build a Facebook over Twitter, but despite Mark Zuckerbergs best efforts you can't build Twitter out of, or in, Facebook.

Facebook is a closed platform that allows some development. Twitter is a very simple API. The simplicity and lack of features "feature" of the service and the fact the Twitter web interface to the service is mediocre has generated a lot of external services and third-party clients.

This has led to the recreation of the "Browser Wars" as "Twitter Client Wars." Tweetdeck vs SeesmicDesktop the first contenders.

Thus Twitter is a protocol and implementation by one company surrounded by an ever enlarging ecosystem which Twitter does not control. They do throttle third-party clients, but this impacts on the user experience and should be avoided. When Twitter flies the Fail Whale, the whole ecosystem dies.

The question then is can Twitter scale. And I believe the answer is no, the answer is federation via an open standard and distributed servers, the way we do email.

Will Twitter survive? Definitely, there will always be a market for their approach combined with mind-share, lack of concern about choices if its good enough and existing network effects will ensure a living can be made. The outer edges of the federation, populated by Laconica installations, Identi.ca being the leading example, along with Leo Laporte's Twit Army among others, will distribute the burden, and to a degree blur the population and services market, but at least it will work.