I'm not sure the MySQL figure is too high.

Compared to Oracle's market cap of $113 billion, MySQL is a relative bargain.

And while their business models are different, Sun is in a position to make a major dent in Oracle's client base.

With Sun behind them, it's easier for big enterprise customers to go along with MySQL instead of Oracle; support costs for using MySQL are also cheaper.

The net effect will be more profits for Sun overall.

Facebook is overvalued. But the numbers 1 and 2 in a market are much more valuable than the rest. If you buy number 1 you are the market leader and if you buy the runner up you can fight for market share or buy some smaller players later on. The number 3 is only interesting if the company can quickly gain a lot of market share. Often those companies are sold off again if that does not happen.

Facebook also has a much better revenue model than providing support for open source software.

Both of the above comments refer to cliches that have been repeated so often that it's tempting to believe them.

We read all the time about number 1 or number 2, but companies survive or die based on profitability and fiscal responsibility, not market position.

Also, revenue model != revenue or, more importantly, earnings. And whether a company has a good revenue model is at best peripherally related to how accurately it's valued.

How do you figure? $3/user/year doesn't knock my socks off.

You are worth your asking price times the probability of getting it. Facebook is not worth $15 bil, not now anyway.