One of the biggest misconceptions in software is that open source equals free. The early commercial open-source vendors like MySQL and JBoss were able to build decent businesses on top of a license/support-only business model, but over time we've seen that approach become difficult to grow beyond a certain threshold.
I suspect that in 2009 it will start becoming clearer as to what you pay for and why you should. Redmonk analyst Michael Cote made the prediction that next year "it will be cool to pay for software" and I agree. It's one thing to consume open-source software and quite another to pay for it.
Most open source vendors have tweaked their business models to include some kind of additional value only available as part of a subscription. This has brought various cries of derision suggesting that the code is no longer good as the community doesn't get to do QA, along with welcoming arms from investors and developers who want to monetize the code.
And it turns out that CTOs want to pay for the additional value, as Matt Asay found out from the New York CTO Club:
From what I heard today, it's a non-issue. Every CTO that spoke up (and it was a very open forum) said that they are happy to pay for proprietary extensions to open-source software, and criticized pure-play open-source vendors for not providing an obvious, compelling reason to pay: proprietary bits. (One actually said that we have built a great financial model...for SIs, not for ourselves.)
Typically we now see an "open core" freely available with "exclusive" or proprietary features only available when you pay. If you are trying to build a commercial business on top of an open source project, this is likely the right answer.
Selling support only has proven to be challenging--and will likely get even harder to do as cash-strapped companies will force their developers to maintain software instead of outlaying more cash.