(1) They say that 78% of the companies surveyed use OSS internally. I'll bet it's more like 78% realize or admit it. It would probably be very hard to run a b

(1) They say that 78% of the companies surveyed use OSS internally. I'll bet it's more like 78% realize or admit it. It would probably be very hard to run a business nowadays without using at least some OSS somewhere. (2) The author's recommended approach of combining some OSS with proprietary enhancements (commonly called "open core") is certainly not the only way to make money from OSS, and I don't think there's even much evidence that it's a particularly effective approach. Some customers are "more than happy" to pay for an "enterprise grade" version. Others aren't, and that number goes up as the price for the proprietary version increases. They'll go it alone, hire consultants, or switch to competitors. A recurring-revenue "subscription" model works surprisingly well. (This is where I should add a disclaimer that I work at Red Hat.) Customers will pay a modest amount, year after year, to get easy/timely updates, priority bug fixes, training, certification, indemnification, access to developers, and so on. People will be "more than happy" to pay for a safety net that they never actually use, and it's easier to keep TCO lower than competitors with this model than with the open core "freemium" model. Less immediate revenue per customer, but more customers and more loyalty. I guess "don't be greedy" is the lesson here. ;)