Edmund J. Walsh and Andrew J. Tibbets on the benefits and risks of Open Source software

Walsh and Tibbets seemed poorly informed about their subject, and as a consequence I think they perpetuated some common misunderstandings about free software:

- I was puzzled by the decision to equate "open source" with GPL-style licenses. This is not how the term is commonly used. Rather, "open source" usually refers to to both GPL-style "copyleft" licenses and permissive BSD-style (what they call "University") licenses. I think tis creates the false impression that most open source software is copyleft, when in reality there are a ton of prominent projects--Apache, BSD, Python--that are licensed under permissive BSD-style licenses. If Walsh and Tibbets want to only talk about copyleft licenses, they should use the generally-accepted term for that concept, or at least make up a world that doesn't already have a different, well-established definition.

- They perpetuated the "infection" theory of GPL compliance. The GPL kicks in when you make a derivative work of a GPLed work. There's a rich debate within the free software community about exactly where the line is, but it's pretty clear that if you (say) build a proprietary application for a GPLed operating system, that doesn't obligated disclosure of the application's source code.

It's worth noting that proprietary software licenses are just as "infectious." Try incorporating a bit of Windows code into your proprietary software and see how far you get. On the other hand, there are a number of "dual-licensed" projects such as MySQL that allow you to pay extra for a non-copyleft license.

- They gave the impression that open source software is at greater risk for patent infringement. The reality is that all software is vulnerable to patent infringement lawsuits, and the licensing terms are simply irrelevant. It's true that if you get your software from a commercial vendor, that vendor is more likely to be sued than you are. But that has nothing to do with open source either--buying from Red Hat gets you the same kind of protection that buying from Microsoft does.

- Finally, I found the discussion of the OIN as an antitrust problem really strange. One of the authors (I wasn't sure which one) said that "if you're thinking only about freedom to operate for the developers, that' a pretty clear, accepted use of a patent pool. But if you get to the point where you say what we really want to do with this collection of patents is sue anybody who asserts that they have intellectual property that is being infringed by our product" then that's an antitrust problem.

This doesn't make any sense. Every large software company amasses a large patent portfolio and "sues anybody who asserts that they have intellectual property that is being infringed by our product." This is because it's almost impossible to create software without infringing some patents. (See Mark Lemley's "Ignoring Patents" for details). There's absolutely nothing sinister about the free software community pursuing the same defensive strategy that proprietary software companies have pursued for 2 decades.<

I don't mean to nitpick, but I wanted to make sure that listeners who are not familiar with free software don't get an unnecessarily negative impression of free software. The undercurrent of Walsh and Tibbets's remarks seemed to be that free software movement is driven an inflexible and radical ideology that is hostile to the interests of for-profit software firms. In fact, lots of companies make good money while using (and contributing to) free software projects. Rather than portraying the GPL as nuisance to be avoided or isolated, I wish they had talked about how companies have worked within the free software community to enhance their profits. Companies that have done this include IBM, Sun, Red Hat, Google, Apple, and many others.