Earth to Bruce Perens: Is anyone listening?
I'm a newcomer to the amorphous debate over open-source software, which probably explains my surprise with Bruce Perens' overheated response to my earlier commentary.
Perens compares himself to Salman Rushdie, whose inclination to stray from orthodoxy made him the target of Iranian assassins. I can't speak for Perens' acquaintances in Redmond, but I can assure him that I mean him no bodily harm.
To be fair, Perens seems to lump me more with Joe McCarthy than with the Ayatollah Khomeini. I am apparently a bigot because I compare devotion to Free Software to religion. And I must also be implying that Perens & Co. are a bunch of commies because I suggest that the General Public License (GPL) will undermine intellectual property rights.
I guess I'll have to live without Perens' approval. But I would appeal to other readers for an objective hearing. The Free Software folks really do think about software in religious terms. As Richard Stallman, the president of the Free Software Foundation, put it, "the proprietary software social system is...antisocial...unethical...simply wrong."
Now, it is certainly possible to like the idea of getting good software without paying for it: Even I enjoy the occasional trip to the free salad bar. And it is certainly possible to be frustrated by the paradox that the price of commercial software is inevitably greater than the cost of licensing another copy, in much the same way that it's frustrating that commercial airline flights take off every day with tens of thousands of empty seats they refuse to give away. But only a true believer would think that businesses are acting unethically when they try to make a profit by charging for software licenses (or airline seats).
By the same token, you don't have to be to the right of Milton Friedman to worry about the impact of the GPL on a system of intellectual property rights that has served the information technology industry pretty well to date. The whole idea behind the GPL (in contrast to other forms of licenses for open-source software) is to deny those who add to or improve software the ability to make a profit on the intellectual property they create.
It is possible that someone will come up with a viable business plan for developing software at no charge and yet still manage to pay the electricity and pizza bills. Who knows: Perhaps Red Hat or one of the other Linux-based start-ups already has. But it would seem altogether prudent to worry that the GPL won't work as well as traditional market incentives in sustaining rapid, effective software development.
While we're on the subject of the GPL, a surprising number of people who ought to know better conflate the GPL with less restrictive open-source software licenses--to assume that whatever is right or wrong about open-source licenses in general is right or wrong about the GPL. I can't speak for Microsoft or anyone else. But I do think the crux of any debate over open source and public policy lies in the distinction.
Much of the software that is distributed with access to the source code puts very few obligations on would-be entrepreneurs. Thus, you can download FreeBSD from the Web, incorporate portions of it in your own software, and sell the package without giving away the source code for your improvements. Somebody may have a beef with this; I don't.
With Linux and other open-source software distributed under the GPL, I am still welcome to download the software, modify it and try to sell the package. But anyone else can try to sell the very same package, since under the GPL I have to make my modified source code available. Very clever, as long as the principle object is to make more software "free" to more people. (Some call this process "viral." But I promise not to use the V-word again, because Perens thinks it breeds guilt by association).
So what's the problem with the GPL? None ordinarily, provided everyone involved understands the rules of the game. My guess is that innovation through software licenses under the GPL will be modest, and that most of the big advances in widely valued software--especially client software, where less sophisticated computer users dominate--will come through traditional commercial channels. But I wouldn't mind being proved wrong.
On the other hand, I do object when Washington effectively takes side by allowing government-underwritten software to be distributed with a GPL (as opposed to a BSD-type) license. That, in fact, is what the Sandia National Laboratories did earlier this month with a software package called Cplant.
Devoted supporters of the GPL obviously get their kicks from occupying the self-defined high moral ground and aren't about to engage in crass debates about incentives. But the rest of us, who see intellectual property rules primarily as a means to an end, ought to be concerned.
The Soviet Union was able to mobilize the resources of a not-very-advanced economy to defeat 200 divisions of the Wehrmacht. But when it came to making shoes and washing machines, socialism proved a bust. Does the analogy apply to non-market methods of developing software in a complex economy, or it is simply another example of my McCarthyite rhetoric? You decide.