All right... the Twitterverse is all up in arms about how crashy it is, and the lack of a business model... well, at least Jake and Radar are... so I figured I'd throw in my 2 cents, and solve both problems at the same time:

How to make Twitter crash less:

1. Ditch Rails.
2. Ditch Ruby.
3. Rewrite it for Python / Django.
4. Use Google App Engine for hosting.

Done and done. Pownce has proven that it way easy to redo everything Twitter did (but better) using Django... and in a remarkable short amount of time. Plus, if you use Django, you can port your entire system to Google App Engine, and get insane scalability and uptime for cheap. Google might even be a willing partner for such a high-profile client with such widely known scalability problems...

I always thought Rails was the wrong tool for Twitter... I'm sure the pragmatic programmers would be all up in arms if Twitter ditched their favorite tool... but who cares? Using the same tool for every job is woefully unpragmatic. "But Rails can do it! Rails can do it!" Ugh... At times like this I let Chris Rock do the talking:

Sure, you can do it, but that doesn't mean it should be done! You can drive your car with your feet if you want to, that doesn't make it a good idea!

Now, regarding the business model, there are these options:

1. Charge $10 per year for people who tweet more than 5 times per day.
2. Engage businesses, sell them "Twitter Appliances," and train them how presence can boost communication and productivity.

Seems pretty damn straightforward to me... at least, that's what I'd do if I had a brand like Twitter. Move into more of an evangelist model, teach people to collaborate with presence, and get into the enterprise before somebody else beats you to the punch. Heck, they could even sell enterprisey books, and be the first "sexy" enterprise app. I'm baffled why they haven't already done so.

In the meantime, I've moved on. Check me out on Friend Feed.

UPDATE: Garrick posted on another twitter business model... The scalability problem is not due to the number of tweets per day, but in the number of followers you have. Some people have thousands of followers, so one tweet per day from a popular person consumes more resources than a friendless one tweeting every hour. Therefore, perhaps you should charge people to be followers? I'm not 100% sold, because that would discourage popularity. Its also vulnerable to Twitter syndicators like FriendFeed... Why should everybody pay $10 to follow Scoble on Twitter? Just follow his FrendFeed instead.