Sometimes you can’t win for trying. Like when you politely ask the kid sitting next to you on the subway who is playing his music a little loud to please turn it down. Just to be contrary, he responds by cranking it up even more.
In my March 15 column, Lowering the Volume in the Software Patent Debate, I suggested with the best of intentions that it might help everyone to lower the volume in the software patent debate.
Naturally, I received some very LOUD responses to that column. Some were coherent and some not even close. So let’s try this again. Instead of YELLING at whomever you disagree with, take that energy and direct it in a more positive way — GET OUT THERE AND DO SOMETHING.
Whenever someone posts a comment about my column, the post often includes a conjecture that I must be on the payroll of some group that wants me to write an advocacy piece favoring them. For the record, I shill for no one. I am not paid to take a particular position, nor have I ever been asked to promote a particular opinion. In any case, I hope that most readers consider the message rather than the messenger.
The debate itself is not clear. Not all participants are advocating positions relative to the same set of issues. Some advocate that open-source licensing should dominate proprietary licensing. Others advocate for proprietary licensing over open-source licensing.
In what should be an entirely separate debate, there are those who argue against patents on software-based inventions and those that argue for patents on software-based inventions. That debate pits one side arguing for open-source development and against patents due to a belief that patents hinder open-source development, against the other side that argues for exactly the opposite.
Some even argue that the debate is not pro-patents versus anti-patents. For example, in Europe, the debate is about the methods of stopping patents on software-based inventions, denying any side that favors patents on software-based inventions.
Why Patents, Open-Source Software Intertwined
The two debates are often intertwined because of the belief that the existence of software patents might wipe out open-source software licensing.
The belief is not unreasonable and proceeds as follows: If an author of a software program releases his or her code under an open-source license and the software is good and useful, it will propagate indefinitely. This benefits the public (who can use this software without having to create it themselves) and the author (who will feel good, who was paid to create it anyway, will get a job because of it, etc.) — until it hits a patent.
Once the patent holder is in the picture, a per-copy royalty might be required, and with many open-source licenses, there is no mechanism to collect royalties. This is fair. Most intellectual property rights only trigger when there is access to the rights owner’s developments, so a software developer can avoid the rights of others by not accessing the protectable property of others.
Not so with patents. One can infringe a patent without ever seeing the patent, so open-source developers cannot ensure that their output is non-infringing.
Real Problem with Software Patents
The real problem is not patents per se, but bad patents. I assume that if someone figured out how to build an anti-gravity vehicle and wanted a royalty for each car developed according to his patent, we’d have no complaints. After all, we didn’t have this before and there is little chance that anyone would have developed it without finding out how from the inventor’s patent application.
Most of the arguments related to so-called “software patents” stem from the apparent ease of invention. I put software patents in quotes because there is no neat category of patents known as “software patents,” just patents that might relate to inventions that could be implemented in software.
The real problem is not patents, but the apparent ease of invention in the software field. If something is easily invented, we seem to think that the inventor is not worthy of any rights to the invention. As far as I know, there are no patent laws anywhere in the world that specify difficulty as a measure of patentability. Instead, most countries look for novelty (you cannot patent something that already existed before you invented it) and inventiveness/nonobviousness (even if it did not already exist, but an obvious variation of it did, you still cannot patent it).
In addition to the notion that one should not reap large rewards for little effort, there is also the notion that patents on software-implemented inventions cause more problems than other patents. Previous programmers might have created the invention and not mentioned it because it was a trivial thing to do. If later programmers came up with the invention on their own, how inventive could it be?
Raise Bar, Lower Volume
There are many examples of patents on software-implemented inventions that should not have slipped past the patent office. This might be due to lack of knowledge of the prior art or too low a test for obviousness.
Software prior art is difficult to obtain relative to prior art for tractors. We know where to find old tractors, but old software can be hidden from view or lost more easily. Most patent offices have historical records of past patents organized by topic, so it is easy to find those that relate to tractors. Not so with software. Software can be harder to organize and there just isn’t that much history yet.
It seems that much of the shouting about software-related patents is based on the perception that such patents are low quality and grant exclusionary rights without inspiring innovation. However, blanket opposition to patents will fail. It has to. There are too many vested interests, and many believe that patents are necessary and beneficial to society in general. They will not go away.
A more promising tack is to call for serious review of inventiveness for patents that do issue for software-implemented inventions. If a software construct is being independently created by thousands of programmers, it is a good indication that the construct is not all that inventive. It is more likely to lead to patent assertions against those with no prior knowledge of the patent, which is likely to lead to the typical anti-patent outrage. However, if some software construct is such that most uses of the construct derive from learning of its invention from an original inventor, less controversy can be expected.
Instead of opposing all patents on software-implemented inventions, how about opposing patents on noninventive concepts? One way to do this is to work on preserving historical records of software developments so that the prior art does not disappear.
Patents should not cover that which a person of ordinary skill in an art would stumble upon when writing code. It should be a little harder to end up with a patentable invention. While that means fewer patents for the less inspired, it also means fewer patent infringements by those that follow.
No matter how good at multi-tasking we are, it is never easy to concentrate with a lot of noise in the background. So let’s all quiet down and get to work.
Phil Albert, a LinuxInsider columnist, is a patent attorney and partner with the San Francisco office of the intellectual property law firm Townsend and Townsend and Crew LLP.
Sometimes you can’t win because of editorial policy (a polite way of saying ‘censorship’ where I live). Why was my comment for Mar 15 not posted? There was no shouting on my part, simply references to sets of papers and journals. Out here in the real world, linuxinsider is seen simply as a mouthpiece for the pro-patent lobby.
Was there really a lot of noise in response to the Mar 15 column? I can see only one comment posted. Or was there? I posted a comment pointing out the factual errors regarding patents, innovation and economics. That is, that there is no positive correlation between patents and innovation or between patents and an improved economy. I referred to academic papers and journals. My comment did not appear. No doubt, the volume is being turned down – but not by software professionals. Only by those deluded into believing software patents are a good thing…
I concur with this comment. I don’t think the volume was too loud the last time. And now it looks like there is really no debate. As yet there are no replys as explaining and/or rebutting the position of those that are against software patents. Being against software patents is a valid explaiable position. All the people speaking against software patents are not all yelling, some of them are saying thoughtful things.
So where’s the debate here?
The problem that I see with the author’s "…opposing patents on noninventive concepts," idea is simply that the patent office isn’t equipped to deal with the question of what’s inventive and noninventive.
So, until it is possible for the patent office to make such determinations with a resonable AM ount of accuracy, I feel that patents on software inventions should not be granted.
The problem really boils down to that fact that, in my opinion, software patents are destructive. That is, they are used to stop competitors and detstroy free markets for software. Most software patents are not really inventions, but features of software that some company wants to own. There’s nothing truly intuitive about "one-click" shopping on the web.
To extrapolate. Imagine if we had always had patents on computing. Someone could have patented the process of putting an impression of a character on a piece of paper, for example. They would essentially owned the process of printing using a computer. Patents like this are what are being created for software, and patents like this are being used to sue competitors out of business. The free market is suffering for this. In my example one company would own printing. Lets just say that their printer was a stupenduously slow daisy wheel printer. No other printers would have come into being, even though we know throug hindsight that there are much better ways to print. Nope, this example would have the patent holder owning the very transmission of a character to a printer, which could have locked every other printer creater out of the business.
Does that example sound too far fetched. Well, what about e-commerce? There are so many patents trying to say that they have a unique way to do x, y, or z. But really its still just a logical extention of the storefront.
Algorithms should not be patentable, because algorithms are just expressions of thought, and while expressions of thought should be copyrightable in their content, patenting the thouoght of one particular function or another is patently ridiculous.
Hmm. Ok. Valid points, however if this:
"(you cannot patent something that already existed before you invented it) and inventiveness/nonobviousness (even if it did not already exist, but an obvious variation of it did, you still cannot patent it)"
is really true, then Microsoft is in some really serious trouble. I would guess that they have patents on a lot of stuff in Windows that also exist on other GUI desktops, some even before MS got involve in such things. Their favorite method of killing competition is taking an ‘existing’ protocol, making a trivial and often unnecessary change to it, then patenting that. Then their is the latest concept of patenting software in combination of hardware, specifically thinks like having one button check your email on a mobile device or similar stupidities, which are hardly new ideas, unless you include the device itself, which is ‘only’ hardware. Patents on such trivial combinations mean that even if you figure out how to still different software on the device, you ‘might’ be breaking a patent by just puting different software on it and anyone that links that one stupid key to email is ‘definitely’ breaking the patent. This is quite insane. I don’t expect anyone to sell me a car with the patent saying I can only use ‘their’ chip in the engine or that if I use a different one I AM not allowed to use the wind shield wipers, because they patented a chip that routes windshield wiper control through the damn chip. Why should I find it reasonable for someone to pull that kind it BS on me with a PDA?
But its not ‘ease’ of creation that is at question here. Its ease of discovery and available functionality. No one can patent a law of physics. Why? Not because someone didn’t discover it first or how to use it, but because attempting to solve a problem using any other method that tries to defy that law it probably going to fail. Maybe rockets could be built to operate from a huge AM gnetic rail gun or a giant water cannon, but the first of those would probably take more power to use than most cities, while the second would probably not even work, so we use rockets. Patent a rocket? Yes. Patent the ‘idea’ of a rocket? That gets to be a bit more problematic. Patent the idea of propelling object using onboard fuel sources and directional thrust (which can even cover the water idea of you pressurize it onboard)? Now you have a *major* problem. This is the situation you get in software quite often. There is only one ‘right’ idea that could work with the resources available and only one or a very small few effective way to do it. Its not about ‘ease’ of creation, its about the impossibility in some cases of finding usable alternative solutions. The more people there are that hoard such things, the fewer there are to improve on or use them and at some point the development of a project could become prohibitively expensive, especially if ‘everyone’ patented, no one gave anything away for free use and ten years from now some new type of compute comes along, most ‘old’ solutions are no longer viable and you have to pay royalties for 5,000 patents, just to turn the damn things on. So yeah, bad patents are the problem, but a bigger problem seems to be recognizing when they are bad. The profiteers will argue that ‘any’ code is patentable, even if its somethings are trivial as clicking a bloody speaker to play music, because they where too cheap to include a synth chip, and the patent office would probably grant it, while completely missing the fact that the same technique has been used since the days of the Apple II. But, they will now argue, this is clicking the speaker on a PDA, not a computer, so it still should be patentable. Right….
As I said, you have a valid point that it is ‘bad’ patents that are a problem, but the people deciding what ‘bad’ constitutes are under staffed and badly trained patent clerks, big corporations, special interest groups funded by big companies and lawyers. Three of the four have no qualifications to make such determinations, though the first of them are supposed to, while the companies have no real incentive to argue against the bad ones, unless they are too small to pay up to those that have patents or hire their own lawyers to fight the same. The privateers have all the ships and their answer to every problem is "Make them pay ransom or walk the plank." This understandibly tends to upset people that can’t do the former and don’t think it should be legal to let companies make them do that later.