GPL v3, Microsoft, Patents, and bloat

• Giant space monster sprouts wings, fangs
• Woman gives birth to MP3s
• Yahoo! asked to explain alleged perjury
• Giant space monster continues its approach
• Yahoos in suits choose cash flow over human rights
• Junior high textbook publishers act like junior high students
• Giant space monster is watching us
• Murdoch conspiracy uncovered
• Web site induces massive loss in worker productivity
• Porn averaged
• Giant space monster threat continues

topics covered:

[PDF version]

I know nobody reads this blog, but as a software and law sorta authority, I feel that I must comment on the GNU Public License (GPL) version 3, even though I'd rather not.

As with inventing a new programming language, inventing a new license creates incompatibilities where there were none before. Whereas before people could just write “This software is licensed under GNU GPL v 2” at the head of their code and move on with the work of writing code, now they have to waste time in what Linus Torvalds calls “office politics.” So what is the gain from throwing out the status quo?

The status quo

To give a real-world metaphor, the Book Thing hands out free books. Just show up on the weekend and take `em. But the inside is stamped NOT FOR RESALE. Hand them out, use them in your school (as many teachers to), but don't sell them for profit. If you find this stipulation to be somehow onerous, then just don't take the book.

And that's about how the copyleft system works. You can download the software and agree to keep it free for future users as well, or you can go buy a copy of Word.

Is this enforceable? As a matter of fact, yes. Haas explains that the current norm, GPL version 2, is enforceable pretty much worldwide. It depends as much on contract law as copyright, but that's perfectly OK, as it meets all the conditions to be enforceable under international contract law conventions.

It's hard for a not-omniscient person to say `never', but I can state with a great degree of certainty that the GPL v2 has never lost in court. It has only been challenged a handful of times, most of which are documented in Haas above. Although written by somebody in Boston with U.S. law in mind, it had a solid victory in Germany when a router company based its software on a GPL project but did not comply with the GPL (translated and annotated version of the ruling).

So we're back where we were: if GPL v 2 has had no problems, why does it need a new version?

Verbosity

To give you an example, let's look at the patent language, which I'll discuss a bit more below. Version 2 specifies that patents and the GPL don't mix, so you have to pick either the free software or patent regime:

[...] if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.

What does GPL v3 say? The same thing, but with many, many more words. Feel free to skim:

Each contributor grants you a non-exclusive, worldwide, royalty-free patent license under the contributor's essential patent claims in its contribution, to make, use, sell, offer for sale, import and otherwise run, modify and propagate the contribution.

For purposes of the following three paragraphs, a “patent license” means a patent license, a covenant not to bring suit for patent infringement, or any other express agreement or commitment, however denominated, not to enforce a patent.

If you convey a covered work, knowingly relying on a patent license, and the Corresponding Source of the work is not available for anyone to copy, free of charge and under the terms of this License, through a publicly available network server or other readily accessible means, then you must either (1) cause the Corresponding Source to be so available, or (2) disclaim the patent license for this particular work, or (3) arrange, in a manner consistent with the requirements of this License, to extend the patent license to downstream recipients. “Knowingly relying” means you have actual knowledge that, but for the patent license, your conveying the covered work in a country, or your recipient's use of the covered work in a country, would infringe one or more identifiable patents in that country that you have reason to believe are valid.

If, pursuant to or in connection with a single transaction or arrangement, you convey, or propagate by procuring conveyance of, a covered work, and grant a patent license providing freedom to use, propagate, modify or convey a specific copy of the covered work to any of the parties receiving the covered work, then the patent license you grant is automatically extended to all recipients of the covered work and works based on it.

I've read it several times, and can see nothing in the new version that was not clearly implied by the first. I can't think of any situation where a GPLed software author would win in court over a patent issue using GPL v3 but not v2--not that any patent issues have ever appeared to date.

Although some lawyers often imply this, more verbose does not necessarily mean more enforceable. Let's say somebody comes up with some creative scheme, like buying CDs from Red Hat and then bundling them for free with their hardware. Under GPL v 2, the activity is still covered, because it's clearly under the intent presented by the word “distribute”. Under GPL v 3, they're in the clear, because if the FSF wanted to restrict that behavior, they would have explicitly stated it with everything else--but they didn't, so they're obviously OK with it.

Judges tend to have a very limited tolerance of word games. If some activity looks like distribution on the face of it, then that's what it is, even if distribution isn't broken down into a thirteen-word categorization in the contract itself. There exists such a thing as a contract that is vague to the point of unenforceable (maybe “satisfaction guaranteed”), but GPL v2 has already won in court and is reasonably clear.

Ad hominem discussion

I could give you a list of similar anecdotes from other people, but there's no need to drive in mean commentary. But do focus on that last comment from Lawyer From Major International Body: the FSF is of decreasing relevance, but its members eagerly want to stay relevant. And thus, we have the GPL v 3.

The FSF made itself a big deal by distributing really good software. Its founder wrote Emacs, a text editor beloved of geeks worldwide, and the FSF still maintains the GNU Compiler Collection, which is the main cornerstone of the entire free software system. I use it daily, and about 100% of my software, from the kernel to OpenOffice to the text editor I'm typing into, was compiled with it. In the 1980s, this was huge, because computer geeks were primarily working on the sort of level where the compiler is visible. But today, few people are that close to the machine. The FSF bitterly insists that a Linux system be referred to as a GNU/Linux system, due to the central position of the GNU compiler and shell and such (why not Linux/GNU?). The fear of becoming a piece of history is almost palpable.

Patents

The first is the text above about the basic question of patent infringement. Version two had this covered: if you refuse to grant a full patent license to any downstream users of the software, then you have to stop using and distributing the work. GPL v3 says the same thing.

The second part of their patent claims is more contentious. wc tells me that it is one sentence which is 13 lines, or 154 words long. The gist is: Remember that Novell-Microsoft deal? Don't do that.

It takes a 154-word sentence to explain that because it was such a specific, odd circumstance. Many members of the FSF saw the move as deeply offensive--I'll admit that I too saw it as shady. But it didn't restrict other Linux distributors in any way--it merely threatened to. But because FSF members took it as a declaration of war, they added a sentence-paragraph to the GPL v3. Without the clause, would such a situation ever arise again? Probably not. Novell took a beating for its deal, and one could argue that their business is worse off for it. The rest of the not-stupid computing world has learned from Novell's error.

The FSF has a more recent argument for the importance of GPL v3: Microsoft has stated that Linux, OpenOffice, and friends violate a few hundred patents held by MSFT. But, a lawyer for the FSF points out, their pact with Novell means that they are distributors of SUSE Linux, and are therefore bound by the terms of the GPL version two or later. When version three comes out, they will be bound to the patent provisions therein. See Groklaw for details on this little story. Notice that our lawyer refers to this scheme as “one more layer of probable defense” against MSFT patent infringement claims. I think he's calling it this because he realizes that MSFT's lawyers can shoot this argument down in time for their second coffee break. I can't explain why he's advancing such a weak legal argument without getting cynical.

But to a great extent, this is all one big misunderstanding. Let me digress a bit to explain why.

Microsoft hates patents

When the ruling in KSR v Teleflex came out, I though I'd be out of business as Mr. Software Patent Expert.

There are a number of problems with patents in the USA today, including:

• Astronomical damages granted at the drop of a hat
• The expansion of what is patentable to include stuff like price lists and equations
• The expansion of US patent law to cover the globe
• Too many stupidly obvious patents

Correspondingly, the nice people at the Supreme Court have heard four cases, whose questions match the above list:

• ebay v MercExchange
• LabCorp v Metabolite
• Microsoft v AT&T
• KSR v Teleflex

ebay, in a 9-0 ruling, curtailed the ability of patent holders to extract damages. Microsoft, in a 7-1 ruling, set tighter borders around patent scope. KSR, another 9-0 ruling from a week or so ago, redefined the means of determining what is patentable. And as for LabCorp, the court decided not to hear it at the last minute. It is obvious to all that software patents would be either curtailed or eliminated in that ruling, and in the ruling of any successor.

But there are people who think that subject matter is irrelevant to the problems of patent law. It's really about details of procedure and the definition of obviousness, they opine, and if we get those things straight, patents for software, business methods, dance moves, storylines, new words, or whatever else will work just fine.

Meanwhile, MSFT is taking billion-dollar losses left and right due to patent litigation. (Yes, billions.) There may have been litigators at MSFT who at one point thought the company could be on top of the patent game, `cause they are the biggest around. But now that they are the target for literally hundreds of stupid patent claims (`cause they are the biggest around), and have had to pay out billions of dollars in a number of settlements, I can't imagine that litigators are still as happy about MSFT's position in the patent game.

Microsoft does have its own patents. If you work at Microsoft and have a patentable idea, they'll give you $1,000 to file the patent I know because I have a pal or two who have taken the bonus. One gave his boss a copy of Math You Can't Use after they finished filing. I am certain that the patent portfolio has paid for itself via various business wranglings like the Novell deal. But having patents does not mean that they're going to get used for suing others. In fact, I know of no litigation to date where Microsoft was the plaintiff in a patent infringement suit. Which is not to say that they never have been, but that they have never been to the scale and extent that it made news, and that I am still not omniscient. Microsoft has never been a patent aggressor--it just keeps threatening to be one.

So what is the business benefit to threatening to sue over Linux right now? It's been about two weeks since their first rattling of the saber, and no new deals have come to light.

But the timing works perfectly with the KSR ruling. All those people who said that patents are safe and perfect now that obviousness has been patched, they suddenly fell silent. Microsoft has shown us that there can still be problems with software patents even after you fix obviousness, damages, and foreign trade issues. MSFT put software patents back on the front page of the Business section.

Which is brilliant. Reverse psychology at its finest. Knowing that they are the hated grownup, they said they love patents simply for the sake of making sure that everybody else would do exactly the opposite, and hate patents. Brilliant, I tell you. And they'll probably ink another deal or two and score a few million in spare cash in the process.

Returning to the FSF, I don't blame our open source lawyers for being worried. It's the lawyer's job to predict the worst case and prepare for it. But it's also the job of the sensationalist to point to the sky and say that it is falling.

Version 2 of the GPL has won every legal challenge it faced. It is immensely popular--partly because a programmer anywhere in the world, with no legal background and a modest command of English, can read it in its entirety and understand its intent. The same goes for any judge. That is, it was elegant and worked well thanks to that elegance. Version 3 follows the unfortunate path of many computer projects past: given a product so good that the users don't need anything further, the producers wonder how they can get out a new product, and so add more and more little details until the elegance is drowned. In short, GPL v 3 is bloatware, that provides little or no benefit over the elegant and effective version 2.

[link] [6 comments]
[Previous entry: "Pricing information"]
[Next entry: "Here comes the ocean, and the global climate change"]

Replies: 6 comments