I would like everyone to watch tridge's talk [.ogv] [mp3] on patents and how engineers can interact efficiently with patent lawyers, to get your knowledge across to them. More formats are linked from End Software Patents, if you prefer audio only.
Focus, please, also on what he's learned about patents, how to invalidate them in various ways, especially the part about proving non-infringement and why that is easier than proving prior art, and then the part about finding a workaround and then publicizing it. He shows how to read a patent filing document. You guys have been effective already in the past in finding prior art, but he is raising the bar with more techniques.
I see some of you are already posting what you believe to be prior art on the Oracle America patents being used against Google, and we might as well do it as well as we can once we are sure Oracle is going to go forward full steam ahead, but there are other arrows to add to our quiver. I'm waiting to see if there is a settlement of some kind, but when I am sure it's going forward, I'll tell you. Patent infringement cases go on for years, usually, so there's time to hope for progress. Larry Ellison is giving a talk at JavaOne September 20 in San Francisco on "Oracle's vision and strategy" for Java. As James Gosling points out, there is still a way to fulfill the 2007 Java pledge by setting up an independent Java foundation, and there are other steps Oracle can still take. In spite of all that is happening, one thing I am clear on. Oracle is not SCO. Oracle is Oracle. Sui generis.
Update: Google announced it will not be attending JavaOne.
But as Eben Moglen
pointed out recently at LinuxCon, the patent crisis in general isn't going away. So it's best that we figure out the very best ways to deal with it. I'm told his talk will be available as video soon, and when it's up, it will be here on the Linux Foundation website.
Linux and FOSS compete on who has the best code, not patent infringement lawsuits, speaking of vision. It's a superior development model. Nobody competes with courtrooms. I'm not saying no one sues. The GPL lawsuits are about copyright infringement, but they are what they say they are, not wolves in granny's cap to fool Red Riding Hood. It's why the code keeps getting better and better.
I've been thinking a lot about that whole issue of using courtrooms instead of competing in the marketplace. I've had occasion to do so, because of reporting on the SCO Group on a daily basis, pretty much, since 2003, among a number of other cases.
And what I have come to believe is this: proprietary software companies in the US, anyway, seem to have been contaminated by Microsoft's behavior and by the general culture these days into acting very much like Microsoft was infamous for doing. That's my analysis. They don't even know how to compete any other way any more, maybe, other than to be as mean as they know how to be. No wonder the economy is struggling. The idea of using, or misusing as I see it, the courts to harm competitors is now not just SCO, and it's a corruption of the system. Patents are a way to make money and worse, they think money is the most important thing, and that destroying the competition is the only way to compete effectively. So they sue even if the patents are simply ridiculous, because it harms a competitor win or lose, just having to deal with the litigation.
It think it's unethical. That's surely not what courts are for. But if money is all you care about, when they start to slip in a market, they often think about "monetizing" the patents.
Have you ever noticed that in any large group of software folks, a mixed group of proprietary and FOSS, the faces of the older guys tell you quite easily who is who? I can usually tell just by their faces whether they are proprietary or FOSS. Older faces tell the tale of who has been naughty or nice.
My mom taught me that when I was a young girl, and I've found it's true. Who you really are eventually shows on your face as you age. When I see the heads of some of the proprietary software companies smile, it looks unnatural, like someone taught them how to look like a nice guy smiling and they do their best imitation, but look at the eyes. Look at the wrinkles. You are what you eat, they say. But really, it's more like you are what you *do*. And it really does show over time what you are inside. When you are twenty, you don't care or you don't think. But as you get older, you find out it really matters, and not just because of how your face looks. Money can't fix you then, not all the money in the world.
All right. Two very different cultures. That's exactly why we here don't like to use their proprietary products and don't recommend them to those we care about -- because if they treat each other the way they do, what will they be willing to do to you and me, mere customers? We can't retaliate or pay them back. We can avoid their products, of course.
As best as I can tell so far, Oracle America is using patents to force Google to use code that isn't as good as the code for such functionality could be, in Google's opinion. And if that's it, and you force people to use inferior code, what really have you accomplished?
But patents will be used by some old-fashioned companies, particularly those that can't compete on merit, just because they can make money that way currently, and while they have been enabled by Bilski to go forward, they also got scared by it into thinking they'd best hurry up and sue, just in case the US Supreme Court decides to toss software patents overboard altogether someday.
And I hope they will. Meanwhile,
tridge expects that patents will be *more* of a problem going forward, and I agree, sadly, that is likely the case, so we need to get more and more effective. How?
The very best is to code a workaround to any patents, if you can find a way. Remember what the Samba guys did with Microsoft? The EU Commission forced Microsoft to provide a bounded list of patents to the Samba team, which right there is better than trying to avoid patents you don't even know about, and they looked at the list and said, no thanks. They didn't need them.
What they achieved in that case is a lot more than that, and I encourage you to reread his description of the process, and if you are rushed, read the part under the heading 'Patents'. Not long after that historic agreement, Microsoft opened up and let the world interact with their protocols more readily. I don't know if they refunded the money
the Protocol Freedom Information Foundation paid them for Samba's and all of our benefit for access to the protocols, but they should have. The point is that you don't want to make a private solution just for yourself. Protect the community at large. It can be done.
The very best technique, if it's possible in a fact pattern you are dealing with, is not only to code a workaround but to publicize it widely, tridge points out. Then the whole world knows there is a workaround, and the patent owner loses the ability to sue anybody with it and get license fees, because all the proprietary victims who otherwise would pay license fees per copy can just use your workaround for free.
That provides a full disincentive to suing FOSS. They'll prefer to sue another proprietary company instead, because they'll just pay up or even if they can code a workaround, they won't share it with the world. Because they are too stupid to share.
That is the perfect counterpunch, tridge says. And as you know, he's put his money where his mouth is, if one thinks a bit about
VFAT, as well as the Samba saga.
The community surely has the skills to make this workaround technique a very effective defense, and where that is impossible to try to prove non-infringement or find prior art. But patents are a minefield, so lawyers are necessary to be effective, as he explains, but while the community lawyers and protective entities figure out a methodology overall, we can right now take step one, and that is to learn how best to communicate with lawyers and to learn as much as we can about the patent system, nauseating as it is.
I respectfully disagree with tridge on one point he made in the Samba article in the following way. I would very much like everyone in the community to get as many patents as you can. If you have a riveting idea, patent it so no one else can do so. Or think about defensive publication. I don't want to just *find* prior art. I'd like to create it. The Software Freedom Law Center and Open Invention Network are there to help the community get it right. OIN would like to hear from you if you get threatened with a patent, by the way. Knocking a patent infringement case out depends on having the precise weapons to do so. You can't fight something with nothing. If they are going to aim patents at you, you can't just stand there and hope for the best. For one thing, patent pledges and promises only cover precisely what they cover, and if they don't cover what you are being sued over, there you are.
The more patents you have, the more likely it will be that you will have just the one you want. But even better is to make it so that they can't patent some functionality because you have it sewn up, even if you never sue anyone. So we can be like epiphytes in the rain forest, growing where we find ourselves, but without depending on the system for nutrients. Avoiding litigation in the first place is always the best solution. And if you have the right patents, the sharks are less likely to choose you to sue. That's talking about companies. What about trolls? If you have the patent already, no patent troll can get it. Trolls don't care about counterclaims, after all, because they have no products. So keeping patents out of their hands is the best mechanism there.
Here's the talk OIN's Keith Bergelt gave at LinuxCon last year on what his organization does and how it works, and at the end he talks about the role you can play.
Update: Here's a TED talk by Peter Donnelly, a statistician at Oxford, an expert in probability theory who applies statistical methods to genetic data, who talks about how some criminal trials have gone the wrong way because none of the lawyers involved nor the experts testifying nor the judge understood statistics well enough to identify serious mistakes that led to wrong verdicts. It's the same principle with tech. If lawyers don't get the tech, and they often don't, how will a judge or a jury get anything related to technology right?