Dana Blankenhorn has been a journalist, writer and part-time futurist for over 30 years.
I have pondered this as well.
FLOSS is often peddled as a "fair" development method which makes software much less expensive.
This is because the savings goes to the customers and not to that big, evil proprietary vendor.
But there actually is a price that's too low, even for the buyers. If the development of the software asset is not sustainable, you are not paying enough. Eventually the software will be abandoned or - as in this case - bought by a company with few intentions of honoring the FOSS ideology.
I find it quite ironic that the old proprietary way of developing software is actually the most fair (socialist if you want): Those benefiting the most will have to pay the most as well.
What Google did with Dalvik was an attempt to reap the tremendous benefit of being a "Java" player - without being willing to contribute back and help sustain the development. Google has indeed benefited a lot from open source. When you are that big you can set up your own support. It is the small and middle-size players who will have to pay for "support" to help sustain the ecosystem.
The FLOSS model is flawed. At least for some types of software.
Sun was not profitable. They never were able to monetize Java. Except for Java ME for which they did receive a fair amount of licence fees from handset vendors. Until Google wiggled a way around that.
I believe that this is also why Microsoft has taken a lot easier on FOSS competition during the later years. They realize that they just have to sit it out. Sustainability of FOSS projects is rare and for a select few. Given enough time most projects will run into the ground or try to cash in selling to a proprietary vendor. The buyer *will* need to monetize the project and from then on the "open" part of the project will become a 2nd class citizen. Given enough time, customers will realize this as well. After the backlash we will still have FLOSS, but then with a more realistic face.
If proprietary companies like Oracle can buy up open source projects and then take back their open source status, how can an enterprise depend on open source software?
It is with the aim of creating this kind of FUD that Oracle has made its moves against Java and OpenSolaris. Most analysts now expect similar moves against mySQL and OpenOffice.
(To the right, a proprietary Dana, not the open source one. All will be explained.)
Why would Oracle destroy assets it just paid good money for?
Two reasons.
- It didn’t really pay good money for them.
- It pushes enterprise buyers away from all open source.
Oracle paid $7.4 billion for Sun. Even with recent drop-offs in revenue (which can be made up) the company should easily clear $2.5 billion in hardware sales this year. If you can buy a viable computer hardware business for three times sales you’re doing pretty well. Oracle got the software assets for free.
Oracle’s moves against high profile open source projects like Java (and presumably OpenOffice and mySQL) also serve as a warning against enterprise dependence on other open source projects.
Open source companies practically live to be taken over. It’s the pot of gold at the end of their rainbow. Oracle is telling their customers that could leave them high and dry.
We assume that the license status of open source is fixed, like baby names. But the status of baby names isn’t fixed at all.
When I was born in 1955, most kids named Dana were boys. By 1970, when the singer Dana Owens (Queen Latifah) was born, most Danas were girls. They will be when my grand-kids are born too. (Picture from Wikipedia.)
What Oracle is demonstrating with Solaris is that such status is not fixed. (Good news for boys named Dana.) Open source can be taken back. Which means that if you come to depend on open source, you may find yourself staring down the barrel of a large bill one of these days, if someone else chooses to be like Larry.
This has always been an Achilles Heel for open source businesses. They can’t make people buy their stuff. They make up for the revenue shortfall with lower sales and distribution costs. But they still have that revenue shortfall.
Thus proprietary companies have a big advantage whenever an open source company wants to cash in. They’re still the most likely buyers. With Solaris Oracle seeks to prove the rights enterprises think they have to open source code can be taken back.
If this can be seen to make financial sense — if Oracle’s strategy is copied even by Oracle — then the corporate way forward for open source may be closed off. Enterprises may decide they have to buy licenses for their own protection, and the open source era ends.
That seems to be the plan, and it’s a cunning one, because as I noted the assets Oracle is playing with were essentially free. All it had to do to get the crown jewels of open source was play a little poker, a game where you win by convincing others your hand is something it’s not.
Open source is not dead, of course. Solaris and the other projects Oracle has bought could be forked. Or can they? Didn’t Google fork Java for Android, and isn’t Oracle now saying in legal papers its control of Java is protected by patents?
And if there’s one thing an enterprise buyer wants to see less than a bill, it’s a lawyer. Lawyers can scare Danas of both sexes.
