Are Software Patents Detrimental to Innovation?
As the holder of sevent patents to date, and notably of the infamous “theme engine” patent, I was interviewed by Chase from osdev.org on the topic. Here’s a reproduction of the interview. The original can be found here.
Chase: Thanks for agreeing to answer some questions.
Arno: You’re welcome.
Would you care to give any background information on yourself or patent 6,188,399? What exactly was your role in U.S. Patent 6,188,399, “Multiple theme engine graphical user interface architecture”?
At the time this patent was filed I was the lead engineer for the Mac OS User Interface. My responsibilities included the Finder and the Toolbox, a part of which was the Appearance manager, which was built to implement interface skinning in Copland.
I am co-inventor of that patent, meaning I was one of the engineers who came up with the clever idea it describes
Looking at your resume I see that you’ve worked in the US, first Apple and now Adobe, but before that you were outside of the US. Most people in IT I’ve talked with outside of the US don’t have nice things to say about computer related patents. What’s your opinion? Good, bad, necessary evil?
I don’t think there is a fundamental problem with the concept of patents. The purpose of patents is to recognize the value of intellectual property. The intent of patents is to encourage inventors to share their discoveries with the world: in exchange for documenting and making available publicly their discoveries, they are granted a protection, limited in time, by the government. It seems like a fair bargain, in principle.
I think there are several points worthy of discussion about software patents, though. First, is the scope of their protection appropriate? While a protection of 20 years may be reasonable for an industrial application, does the same scope make sense in a field that is barely 50 years old and evolving so quickly? One can reasonably argue that a shorter protection period would be more appropriate.
Second, is the patent review process effective? To be valid a patent has to be useful, novel and non-obvious “to someone skilled in the arts”. That is, it has to be a genuine invention, something that no one ever thought of doing before, and something that a reasonably skilled engineer would not come up on his own when faced with a similar problem. Unfortunately, due to the volume of patents application being produced nowadays, it has become more and more difficult for patent reviewers to do their job as thoroughly as they would like. I believe that quite a few questionable software patents have been granted (but none of mine, of course ) either because they are in fact obvious or there is prior art available, and therefore they are not novel. Unfortunately, that only gets determined if someone bothers to challenge a patent. On the other hand, if you are confident that a patent is invalid, particularly if you know of prior art, you may consider ignoring the patent and waiting for the patent owner to contact you to present your prior art. A patent can be invalidated after it has been granted. Note that I am not qualified to give legal advice
Do you think other countries will adopt similar laws?
Many countries already have similar laws. The protection of intellectual property is important in many parts of the world and is becoming increasingly important. Without this kind of protection the incentive to invent would be reduced. The small inventor who comes up with a better encryption algorithm and publishes it would see his work ripped off by larger companies without any need for them to compensate him. On the other hand, larger companies would have no incentive to disclose their intellectual property, but instead would keep their inventions as proprietary information they would not share with anyone else, gaining in perpetuity exclusive ownership of some knowledge that could become fundamental to our society. There are some discussions about what can be patented and how to distinguish between inventions and discoveries. For example, mathematical formulas typically cannot be protected by a patent. Software patents sometimes fall in a grey area.
One of the complaints with computer software patents that always gets mentioned is that the “invention” is not a new concept. Not only are there always mentions of prior art but many technically savy people would consider many patents as obvious evolutions. Do you think there is a problem with the review process?
Yes, I think the patent review process could be improved. I think that the inventors and the public would benefit from a better review process. In my experience, when submitting a patent one of the difficult part of the application process is researching prior art. You want to avoid the expense of submitting a patent application that is going to be rejected or to rely on a patent that could be declared invalid after being granted because there is some prior art that was not discovered at the time of the submission or the grant.
Similarly, during the review process, the reviewer wants to do the best job they can to ensure the idea is original and non-obvious. Unfortunately, there is no convenient process to research prior art today.
I think an organized community-based process would be welcome and helpful. Imagine that you are a reviewer about to review a patent application. What if you could post a query to a site asking for prior art on an invention to which anyone could respond to with citation or evidence of prior art? Sometimes the IT community does band together when aggressive companies start to issue threats to enforce crucial patents such when Forgent asserted claims that impacted JPEG. Maybe we could generalize this?
As an engineering manager, I’m often responsible for reviewing our work and selecting what could be patentable. I have to consider both the “obviousness” of the invention, as well as any prior art I know of. Only if I find something that passes both criteria do I involve our patent group. If they agree with me, with then proceed with an internal review. If the internal consensus from other engineers in the company is that the idea is non-obvious and original, then we involve our patent lawyers at which point we submit an application to the patent office. So, we try to do what we can to limit and eliminate frivolous patents, but we may not always be successful.
When this patent was being mentioned on IT news sites there were a few responses that went something like “What?!?! They patented THEMES???” Even more now when almost all computer software can be themed the idea that theming is patentable seems bizarre. Of course everything seems obvious when looking back in time. During your time at Apple was patent 6,188,399 seen more as a significant enhancement to an existing invention or something new?
LOL Yes, I remember reading some of the reaction when this patent was discussed. I was wondering how long it would take for someone to ask me about it. Apparently, the answer is 5 years!
Patents are legal documents. As a result, they are very precise but they are also somewhat difficult to read. Not too many people actually took the time to read this patent and understand what it applied to.
This patent doesn’t cover the concept of themes or skinning. First, that would hardly be original and second that would actually not be something that can be protected by a patent. A patent cannot be used to protect an idea or concept. It has to protect a specific manner to accomplish something. In this case, this patent is fairly narrow and applies to procedural skinning, if you will. That is, the ability skin or theme software using not just images assembled together (I’m not aware of any patents on that and I think one could argue easily (a) there is a lot of prior art on that and (b) it is obvious to a skilled engineer) but using procedural code to specify a skin or theme.
When we came up with this idea back in 1997, first we had a “hey, that’s a cool idea” moment, which is the first sign that something might be patentable then a “hey, I’ve never heard of anyone doing anything like this before”, this being the second sign that something might be patentable. We then put in motion the process to apply for a patent, which included doing research on prior art, describing it, etc…
So, the concept of skinning wasn’t new (although it certainly wasn’t as widespread as it is today), but how we wanted to do skinning was.
In terms of IT patent news this patent is rather old, most mentions were back in 2001 with articles such as http://www.theregister.co.uk/2001/03/09/apple_patents_desktop_themes however there have still been mentions of this patent over the years (mostly on KDE or Gnome mailing lists). The majority of articles back in 2001 all seem to suggest there is prior art. To your knowledge have there been any legal challenges to the patent?
I’m not aware of any legal challenge to this patent. However, like I said this patent is frequently assumed to being much broader than its actual scope. I don’t believe there is any prior art that I know of, but I’m open to being proven wrong. Also, contrary to what the Register states, Themes.org does not constitute prior art, since again the patent doesn’t cover the concept of skinning *and* the patent pre-dates Themes.org (I think in fact we inspired Themes.org).
After browsing through the U.S. Patent office website and even the patents related to patent 6,188,399 I can see why some people say that almost everything you can do with a computer has been patented. For those of us in the U.S. interested in creating operating systems it seems the entire process is basically illegal or at the very least could get us sued back to the stone age.
I’m not sure what you mean by illegal. By definition itself, patent protection is a legal process. You might think it is unfair, or non productive and that the law should be changed, but I don’t think you can call it illegal. I also think that there are still many new inventions to be made, otherwise I don’t think I would still be working in this industry.
Do you think amateur operating system developers really have anything to worry about? No, I don’t think so. If you don’t intentionally violate a patent, I wouldn’t worry about it too much. Patents are designed to protect the public interest. They are not granted for the protection of big companies, in fact just the opposite.
Patent infringement cases typically involve willful infringement, that is someone who knew what they were doing when they decided to use the intellectual property described in a patent. Typically, if you are unknowingly infringing a patent you are given the chance to rectify by using some other mean to achieve the desired effect. So, you would not be prevented from skinning your app or your OS, as long as you did it some other way than the method described in the infamous patent above.
Again, it’s not the idea that’s protected, it’s the specific method to implement the idea. So, you could either choose to do it using one of the “obvious” methods for which there is ample prior art, and which are therefore not protected by patents *or* if you find there is so much value in the particular method that you absolutely want to use it, but yet cannot find prior art about it, you must recognize that the idea is original and valuable and you should give a tip of the hat to the inventors, maybe by offering them to license their patent.
Have you heard of the SCO vs IBM patent case involving Linux? Do you have an opinion?
Actually, the SCO vs. IBM case is not a patent case. SCO had various claims against IBM, but none involved violating a patent. On the other hand, IBM counter-sued SCO claiming infringement by SCO of some IBM patents. So one could argue in this case that the patent system works since IBM is using it to defend itself against a predatory company
I was willing to give SCO the benefit of the doubt regarding the merit of their claims and assume that their actions were really just a sincere mean to defend their intellectual property. However, their behavior leads me to think that SCO is behaving as a predatory company or at least as a desperate company trying to sue themselves out of bankruptcy. Personally, I’m rooting for IBM on this one
Do you think that companies such as Microsoft, IBM, and Apple apply for so many patents just to acquire an arsenal of patents? To some degree it seems that once a company acquires enough patents it becomes safe to function. A game of thermonuclear war if you will.
Large software companies file patents to protect themselves from being sued by a smaller predatory companies (such as SCO). I am not aware of any instance where Microsoft, IBM or Apple have gone after a small (or large) inventor and sued them for patent infringement. The business models of these companies is to innovate and sell (and protect) their innovations, not to write patents and license them.
Very often large companies negotiate with each other cross-licensing agreements, meaning that they pool their patents together. Companies therefore have an incentive to accumulate patents, as the set of patents they have is considered a valuable commodity in those cross-licensing deals. And the reason they are interested in those cross-licensing deals is to cover themselves, not to go after others.
A few firms have a business model based solely around the acquisition and aggressive enforcement of patents (Forgent comes to mind). I personally don’t find those companies to be providing a very valuable service to society. I see them as an unavoidable consequence of our system but find what they’re doing to be morally objectionable, although legal. I don’t think they contribute to making the world a better place.
Without worrying about your current employment contract, if you had an original idea involving computer software would you personally patent it? Why?
I would consider it, certainly. However, patenting an original idea is not always the best course of action. I’ve also worked in academia where you publish and share your results for fame and no profit. It’s a different kind of reward I’ve also been involved with some startups that did have valuable intellectual property but decided it was in their best interest to keep it as proprietary information. They basically did not want their (potential) competitors to know what they were up to. unfortunately, some of them folded and as a results their ideas went with them and were never shared with the world. They were good ideas too. Maybe someone will rediscover them someday.
Thanks for the thoughtful questions and for not assuming I was evil for having my name on a patent!