Put three geeks in a room and it won't take long to start an argument. Well, analyst Dennis Byron, veteran open source exec Stuart Cohen, and ex-Microsoft developer Keith Curtis weren't exactly in the same room, but all three have provocative opinions about the future of software in general and of open source in particular.
Byron, who has 30 years of technology experience under his belt, thinks that the open source era is coming to an inglorious end, with Linux and the like becoming "an asterisk in the history and future of technology." Moreover, he claims that the long struggle against Microsoft's undue influence hurt the software industry more than it hurt Gates & Co.
[ Keep up with open source news and views with InfoWorld's Open Sources blog. ]
Curtis, whose opinions are contained in his book After the Software Wars, now available on the Web, says his former employer is "toast" and believes that free software is the key to breathtaking technological progress in everything from space travel and artificial intelligence to automobiles that drive themselves.
Cohen, CEO of Collaborative Software Initiative, argues that open source's greatest strength -- its top-notch code -- is also its greatest weakness. "Open source code is generally great code, not requiring much support. So open source companies that rely on support and service alone are not long for this world," he wrote recently in a BusinessWeek blog.
Why should you care? It's simple. Bad times -- and who can doubt that they're here -- are the most important times to question our basic assumptions. I'm not at all sure that any of these gentlemen has it right, but they certainly raise issues that developers, managers, and investors in open source should be thinking about.
(I had a chance to speak to Byron and Cohen after reading their blog posts, but I wasn't able to reach Curtis, so I'll simply quote from After the Software Wars.)
If It Ain't Broken, Why Pay to Fix It?
When Cohen's post first appeared, he took some heat from people who thought he was saying that open source is sick. So he was quick to tell me that he was misunderstood: "Two issues get blurred. The health of open source software has never been better when we are talking about the quality of the code. The number of people using it and the number of enterprises deploying it have never been higher."
But he maintains that the classic open source business model no longer works very well, and so it has to change.
If open source code is so solid that it doesn't require much support (his assumption, not mine), how then does an open source company make any money? There are basically two ways, argues Cohen.
The first: Get small. He figures that companies can live on lower revenue and margins by not spending much money. The downside: Unless the company has invented something so terrific that it'd be a takeover target, investors will have little interest. The other road: Emulate Red Hat.
"Red Hat, arguably the most successful open source company, has also found ways to add value beyond supporting the Linux kernel. It adds substantial layers of software on top of the kernel, a solid piece of software that needs little support, in order to provide additional value to its customers. If Red Hat relied on supporting the Linux kernel, it would go out of business simply because the code is so sound," Cohen says.
Collaboration Via Open Source Can Reap Big Savings
Cohen's Collaborative Software Initiative offers a different model. CSI pulls together teams from companies interested in developing applications that don't enable competitive advantage. CSI's developers do the heavy programming work in collaboration with subject experts from the participating companies.
Interestingly, some recent work by Dennis Byron, who has no relationship with CSI, sheds some light on the advantage of that model. An analyst with IT Investment Research, Byron looked at a study by the Linux Foundation that claims that the total cost of development of Fedora 9 would have been $10.8 billion if it had been done from scratch by one company using a closed source development methodology.
Byron applied the Linux Foundation's rather academic methodology to the development costs of Microsoft's Vista and closely related products and found that Redmond's costs were approximately $14 billion, which lines up pretty well with the Fedora costs. The real takeaway from Byron's exercise is that the Linux Foundation estimate is a reasonable one to use, and that the collaborative nature of open-source development is a good deal for investors because cheaper is obviously better if quality is consistent.