Weakening Copyright is Weakening Free Software (Part 2, Copyleft; Embrace/Extend)

In Part 1, I quickly reviewed what copyright is and how the GNU General Public License (the GPL) uses copyright law to allow the original author of a piece of software to make it available for widespread use while simultaneously ensuring that they (and everyone else) receive the benefit of any changes that other people subsequently make to it.

For and Against: Copyright Law and Copyleft

Many people advocate reforming copyright law, many of them from the software industry and related fields. And, depending on the reform, it would be fair to count me as one of them. There is also a vocal group who believe that copyright is anachronistic (digital copies have virtually no cost) or contrary to the social good (the free-flow of information is more important than ever in today’s society). Anti-copyright advocates will often use the slogan “information wants to be free,” and maintain that attempts to enforce copyright law are counter to the flow of progress and doomed to failure.

As you may have guessed, I don’t agree with the copyright abolitionists. I also don’t agree with a related set of people, mostly programmers, who are in favor of Open Source software, but feel that the copyleft provisions of the GPL are (somehow) “wrong.” Given the subset of these programmers I’ve known personally, I think that for the majority of them, the objections are relatively shallow: the fact that some software uses a copyleft license imposes extra work on them (making their modified version available publicly) that they’d rather not do, or forces them to talk to their employer’s legal department, or prevents them from using the software entirely because of their company’s policies/decisions. I’ve met virtually no developers who are self-employed, or not employed as programmers at all, who have these objections.

The rest of the anti-copyleft programmers I’ve known and read make arguments in the same famillies as the justifications for copyright law: moral and pragmatic/economic. The moral/ethical arguments against copyleft are much like those of the copyright abolitionists: “software wants to be free” and that it’s unreasonable to call something completely intangible “property.” Somehow, though, they feel that these arguments apply to software more than they do to other copyrighted works. Perhaps they see software more as a collection of useful parts than as an integrated whole like a novel or song. (Although I would imagine that sampling musicians would disagree.)

One interesting anti-copyleft argument is that the GPL (and similar licenses) are not “free” because they deny other people the “freedom” to keep their own modifications to a software package private/proprietary. At first, this argument might be appealing, but it is essentially analogous to saying that a tolerant person must also tolerate someone else’s intolerance. These people are worried about their “freedoms” with respect to something that first had to be created by somebody else. And even though they don’t have to sign a document before copying the original author’s work, the license is still a contract between the two. And if they don’t want to have their “freedom” somehow “taken away” by that contract, all they have to do is not copy the other person’s work in the first place.

Some people who complain about the GPL taking away their freedoms subscribe to the anti-copyright argument that an unsanctioned copy doesn’t take anything “real” away from the original author. But to them I say “How can you complain about the GPL requiring you to let other people copy your work if you believe all restrictions on copying are wrong?” At which point, we arrive at the root of the matter. The people concerned over the GPL’s “taking of their freedom,” underneath all of the moral and economic arguments, actual fall into two groups:

• the lazy: they don’t actually object to other people copying their work, they object to having to go to all the trouble to publish their work for others. Yet they want to benefit from the (presumably much greater) work of original authors who did publish.
• the intellectually dishonest: they’re working in a compiled programming language, so they know that even if copyright were abolished, as long as they release only executable binaries, nobody else can build on their work, even though they’re building on the original author who published their code as source.

Finally, the argument against “depriving people of freedom” cuts both ways. If the copyright abolitionists succeed (either completely, or by carving out some kind of software-specific exception), then they will have deprived me of my “freedom” to only share software source code with those who will share with me in return.

Embrace, Extend, Extinguish

“Embrace, Extend, and Extinguish” is the alleged internal name of Microsoft’s strategy for moving existing standards and technology from being competitive to being Microsoft-propriety. The method is practiced as follows: (1) wait for somebody else to implement something cool, (2) implement a clone of it yourself, (3) add a handful of minor tweaks so that your clone isn’t compatible with the original, (4) use your existing dominance to bootstrap your clone into the marketplace. After step 4, Microsoft’s market share grows monotonically barring a major success by a competitor, because each customer that changes to the MS-specific version creates for themselves a significant cost to ever change (back) to a competing version.

Clearly, the existence of software licensed using copyleft in a particular technical niche doesn’t prevent Embrace&Extend entirely. Any organization large enough to practice this strategy is large enough to implement a clone entirely from scratch. However, having public, non-copyleft sources certainly gives the “extender” a leg up; even more so if the public version of the technology continues to be maintained. Without a copyleft license, and the underlying copyright law to enforce it, the existence of public sources for a technology is an even greater boon to predators and exploiters than it is to the cooperative community. Who knows, if the GNU software wasn’t licensed under GPL, perhaps Microsoft would have started using it as the basis for Windows in the late ’80s or early ’90s?

The article Why I Don’t Use the GPL asks the question “what harm is really done [if a company uses non-copyleft source code] in their product?” The author concludes that there isn’t any harm worth considering (“The open-source project still exists.”), and that there would be benefits to the community-at-large if proprietary software contained larger fractions of open-source code. But in the end, I think that this is a naive analysis. Perhaps it would be accurate if the biggest software companies were what we now consider mid-size, or if they were less likely to look for opportunities to gain advantage exploitatively.

This concludes in Part 3, with a discussion of the different types of Open Source licensing, and what the open-source world might look like in the absence of copyright law.