Software Patents
I heard on twitter that the patent office is currently soliciting public feedback regarding software patents following Bilski v. Kappos. http://www.fsf.org/news/uspto-bilski-guidance. The following is the letter I sent in response. If you agree let them know. The deadline is tomorrow. Bilski_Guidance@uspto.gov
I have been a software developer professionally for more than 10 years and a hobbyist for another 10 or so years before that. I strongly believe that software innovation proceeds only in spite of patents not because of them.
There are a least a half dozen patents on well established computer science, for example the doubly linked list has somehow been patented (http://www.feld.com/wp/archives/2010/03/the-doubly-linked-list-appears-to-have-been-patented.html) despite that it has been an essential data structure taught in every computer science class since forever. There are too many of these type of patents that it is most-likely impossible to write any software that does not violate a least a half dozen software patents. It is a bit like being an artist painting a on canvas where every type of brush stroke has been patented. If you do manage to be successful and a patent troll comes knocking then maybe you can settle, however if you do not even start to paint then you have no chance of being a successful artist.
My hope however is that many of these patents would not hold up in court and that these patents are primarily just for show. It has become essential for companies to have numerous patents in order to protect themselves, not from their ideas being copied…. but rather as a form of currency to trade in the event of future lawsuit. “You claim I violated 2 of your patents. Well look here. You violated 5 of mine! Can we settle?” Even if many of these patents could probably be disputed, the cost of doing may be too high, and large companies will often create patent agreements with their competitors. In this way patents are more like nuclear weapons that companies will stockpile for defense, but the only ones on which you can use them are the little guys who don’t already have them.
The thing I find most disturbing is how many innovations arise within the open software community are stolen by commercial companies renamed, re-branded, and patented. This makes prior art searches more difficult. And even if found, psychologically, the name change somehow makes any trivial difference seem somewhat more novel. Unfortunately, most open source developers do not get paid for there contributions and often do not have the funding to protect their innovations from theft. Additionally, the closed source nature of most software companies makes it difficult to determine or prove that one’s copyright has been violated. However, many open source developers want their ideas to be used by others. That is a big part of why they made the source open in the first place, but when companies later patent these ideas as their own the open source community is disheartened.
I am currently working for a startup, which has been granted several patents. Although the patents are supposed to help protect are the company, I personally think most are not novel enough to warrant 20 year patent protection. Still, none of these patents capture the essence of what our company is about, and the thing that will make or break our company is NOT the software PRODUCT, but the entirety of the SERVICE we provide.
I am not entirely opposed to short term software patents. However, 20 years is WAY WAY WAY to long for a software patent. If software patents are continued, the time frame must be substantially reduced to better reflect the actual time frame in which the return on investment is realized. Software is now being created at breathtaking speed, what once took several developers years, can be done by a couple college kids in months. 3-5 years should be more then enough time to develop and prove an idea is or is not profitable and even then that is probably still too long.
Patents mostly just enable large companies to stagnate and stop innovating. I have seen a lot of crappy software and companies that have stopped supporting there customers and failed to respond to customers changing needs. After a few years their customers are left pining for alternatives, alternatives that cannot easily be created without risking a patent lawsuit.
Usability and simplicity are extremely important to most consumers, how the idea is presented is just as important to a product’s success as the idea itself (just ask Apple). If our implementation is crap and the user’s don’t use it then we should fail and we do not deserve to have a monopoly on the ideas and methods used to provide that service. Allowing companies to patent the methods and ideas used within software is a bit like allowing a lemonade company to patent all the methods of extraction and measurement of lemon juice. The quantity and quality of lemon juice, water, sugar, is far more important to the end result than the methods used to produce it. Although idea of lemonade may have at one time been novel, when the initial producer stops caring and shifts to lower quality ingredients and waters down the product this should be an opportunity for someone else to step in and produce a better product! If software is to be protected as a product then only the end result itself as sold to consumers should be protected in its entirety, not the individual ideas, artifacts, and methods used in its production.
In summary, the patent system currently does more harm then good. Patents today are granted for way too many things which are neither novel nor new. Rather than protect startups, they enable big companies to form alliances that shut out competition. They allow companies to stagnate rather than innovate, and fail to capture the real value of the service provided. If the software patents are allowed to continue, they must be limited to the final artifacts in their entirety and the time frame of validity must be substantially reduced.
