On The Evilness (Or Not) Of Software Patents
from the ignoring-the-elephant-in-the-room dept
Dosquatch writes in to point out Paul Graham’s latest essay asking if software patents are evil. For those of you used to the arguments on this topic there isn’t that much new — but some of it is framed in a different manner. His basic argument actually seems to contradict itself — and misses the biggest issue of all. He starts out by noting that software patents really shouldn’t be seen as being any different than other patents — though, he doesn’t support that argument very strongly. It ignores, for instance, the fact that software is already protected by copyright. However, he says that the real problem is that the USPTO patents too many things, because they just don’t understand software. Those aren’t new arguments. From there, he goes on to say that, given all that, the furor over patents is no big deal and that they don’t really hinder or help innovation in either direction. Instead, he notes that they’re useful for nuclear stockpiling purposes, so that if someone attacks you on patent grounds, you attack them back. He admits that patent hoarders or “patent trolls” are a problem, simply because you can’t sue them back in the mutually assured destruction game — because they don’t make anything. But, adding it all up, he doesn’t think that patent really matter, other than for those defensive reasons, or during an acquisition when a company wants something clear they can say they’re getting for their money.
What he’s missing in all of this is the money factor. A lot of money goes into filing for patents and in the whole game of fighting back against those who sue you over patents. While Graham claims that the only companies who get sued for patent infringement are successful and that the innovation has “already happened,” that ignores the fact that innovation should be an ongoing process — and when money has to be spent on lawyers to fight off patent suits, it’s not going back into continuing innovation. So, it may not stop innovation, but it can certainly slow down the rate of innovation. Considering that the real trick to innovation is having continual fleeting competitive advantages, slowing down the rate of innovation can be quite harmful. It may “even out” among all the companies competing in a space (though, it doesn’t seem like that’s actually true when you look at the evidence), but there’s still a tremendous waste product in terms of money going to lawyers that should be going to innovating and bringing new products to market. To Graham, patents are simply a reality of the software business — and if you’re in the business, you have to play the game. That’s certainly true. That doesn’t, however, mean people shouldn’t look to make things better.
Comments on “On The Evilness (Or Not) Of Software Patents”
Nature of progress
Innovation is cumulative. Some types of innovation build on what has come before and in turn enable further innovation. Patents can slow down that further innovation significantly.
Watt’s patent on steam engines stopped development of improvements for almost thirty years and the Wright brothers’ patent resulted in the US ceding leadership in aeronautics to the French.
Requirements for holding a patent
I believe that one of the requirements for holding a patent should be that you have a VIABLE product that you are currently offering that includes the patented material, not just the patent itself. That would get rid of the patent trolls/hoarders because they would not have a viable product based on the patented idea. That would also get rid of all the stale patents that are not being actually used for anything other than leverage against someone else.
Some Software ideas are worthy of patents
There are a few things in the software realm that are worthy of patents: RSA encryption comes to mind.
The purpose of a patent is to protect someone who spends time, energy, and money developing an idea/product and having someone else swoop in, reverse engineer and make all the money on it. In this way, software patents are the same as hardware or mechanical patents.
The problem really is that patent clerks and patent lawyers do not understand software. What is intuitively obvious to a software developer may not be obvious to anyone else — and hence, a lot of bogus patents are issued.
Both sides have compelling arguments(the sane arguments anyhow).
I think that a system that can satisfiy both sides at least to some degree would be the system that has been imposed on the drug companies in many countries.
The inventor of the software recieves a “lock” patent on the new product for a limited time, lets say 10 years, then it becomes public domain and developers can userp the code for marketing their own brand of the software, with or without changes.
It works in the drug business and has for many years, at least in Canada and has not (look at the market) seemed to injure the original developers of the original drugs.
Of course there will be many arguments if such a method would be implimented, but the first thing that I think must be done is that the patent offices must come to grips with the fact that there are not enough personel with the qualifications to evaluate what should or should not recieve patent protection in the first place.
Patents should be for innovation only !!
I have to agree with Howard, very little software is worthy of patents.
Most ideas for software come from polls and public demand so the idea can not be considered ‘uniquely innovative’ since almost everyone has thought of it independantly.
Secondly, software development is more like carpendry than artistry. Given a discription of what to produce any competent programmer can produce a working program using standard design and construction practices. No innovation nessessary!
Another problem not mentioned is that alot of software can only be produced after the hardware becomes available to make it useable. Hardware is developed secretly and disclosed privatly to developers who rush to patent every possible application of the new hardware before the general public is made aware if its existance. On wallstreet they call that insider trading and it will land you in jail, but in the technical world it’s the status quo.
As for solutions, they first need to issue fewer patents and secondly patents need to expire in a reasonable time frame, perhaps based on market saturation. Record albums go ‘gold’ and ‘platinum’ based on sales and market saturation. Similar saturation points could apply to software when it becomes widely adopted enough to be considered a standard, then it is no longer in the best interest to allow someone to maintain a patent monopoly on it. Other arrangements would have to be made to deal with patent trolls and the use of patents to prevent valid competition. Perhaps a forced licenseability of patents not shown to be in use…
at any rate … there is a lot of work needed to fix the patent system as it stands today. And the new patent initiative being taken by software companys could kill the industry completly if something isn’t done soon.