Tuesday, June 27, 2006

It's Obvious...

I usually steer clear of controversial subjects such as the notion of software patents, mainly because I am somewhat conflicted on the subject myself since I've filed for patents here at Borland.  On one hand, one should be able to gain patent protection for a new and unique invention (or in the case of software, a process).  On the other hand, with the seemingly whacky patents that have been granted over the years for what many agree are obvious uses of existing technology (can you say, OneClick?).  Well it seems that the U.S. Supreme Court is actually going to hear a case and possibly rule on what should be considered "obvious to a person of ordinary skill in the art."  This article on ArsTechnica outlines this case.  There's also some links to the actual petition.  What is really interesting in this case is that the petioner actually won their case in the U.S. Court of Appeals!  However they felt that the court didn't rule on the premise of their argument which is based on a 1952 federal law regarding the obviousness of an idea or invention.

Regardless of what side one falls on the whole notion of whether or not software should be patented, the fact of the matter is that software is patented.  So if you refuse to file for patent protection on your new and unique ideas simply on principle, you may find that that is all you've got to fall back on when someone becomes wildly successful based on your idea.  Or even worse, comes after you with an infringement suit because they patented the idea.  Until the current U.S. patent system is reformed or at least clarified, your only choice is to take a defensive stance and try and patent as much as you can.  In many cases the best you can hope for is a stalement if someone comes after you with an infringement suit.  If you have your own portfolio of patents, chances are the person filing suit against you may be infringing on one of yours.  So the only recourse you have is to cross-license the patents.  If you had nothing in your portfolio, then you could be liable for major damages including on-going license fees.  The really nasty kicker is that if the plaintiff can show that you had prior knowledge of a particular patent and infringed anyway, you could be liable for triple damages!

Whether or not this case presented to the U.S. Supreme Court has any affect on the patent system remains to be seen.  It does, however, look to be a step in the right direction.

7 comments:

  1. Actually, software is NOT legally patentable. It explicitly falls outside the described boundaries for what is patentable.


    This does not prevent the patent office from selling patents or the courts from trying to enforce those patents (the courts could care less about the validity of the patent, just its existance deeming it not their role to police the patent system).


    So, how do you stop the patent system from selling patents on things that it should not? How do you repeal those patents it should not have issued?


    Frankly, I doubt we'll ever see the system change. Precident has already been well set and the assumption that software can be patented has become fact. Now, everyone can apply for patents, no matter how lame and unoriginal and expect to get it.


    Take for example, a patent that was issued on function pointers. This higher level language concept LITTERALLY maps to single machine code instructions, and is an obvious extension of indirect calls, jump tables and interrupts. This didn't stop the patent holder from getting the patent, and it did not stop them from licensing it to other companies.


    Is it ethical to exploit the ignorance of the system like this? No. Is it moral? No. Is it legal? Apparently.


    Will the supreme court do the right thing and stop insanity like this? No. They simply don't know the field well enough and unless you could appoint programmer judges, how would they have any real clue what is actually obvious?


    No, we are all going to have to wait for the patent system to completely and utterly collapse under its own weight. After that, no one will have patents of any kind. This will remove patent protection from those whom the system was intended for just because patents were issued to those for whom the system was NEVER intended for.


    And none of us are better off for any of it.

    ReplyDelete
  2. Allen,

    Mutual destruction only works when the other party has a stake in the software industry. When companies exist soley to buy/trade software patents, there is no chance you can threaten a counter suite to make the problem go away. You will end up paying your attorney's or the extortionist.


    IIRC, software is the only thing protected by both copyright law and patent law. One reform for software patents could be to require a working sample of code that demonstrates the patent. This sample then enters the public domain.

    ReplyDelete
  3. Copyright protects only the *expression* not the idea. Patents protect the invention or the process.


    As a matter of fact, it is somewhat true that software is not patented, per se. It is the *process* implemented by the software that is actually patented. The actual implementation of that patent is what is protected by copyright. So you can infringe a patent but not copyright on the same process.


    If you want to get into what silly things the PTO allows to be patented, look at gene sequence patenting.

    ReplyDelete
  4. As you say it is a step in the right direction, but unless and until some way is found to provide the patent office with the domain knowledge needed for any particualr industry, there will continue to be silly patents passed.


    I don't think software patents, per se, are illegitimate, the problem is simply expertise in judging against proper patent criteria (both obviousness and finding prior art).


    A good example (IMO) is a specific encryption algorithm, say AES. A patent for that is perfectly reasonable because it only protects that particular process, but does not stop others from developing competing methods (e.g. BlowFish). The problem currently, is it would not be hard to pass a patent that covered ALL encryption methods - i.e. patenting the *idea* of encryption rather than a particular method of doing it.

    ReplyDelete
  5. Allen, with all due respect, and much is due, you have to have a very diabolic sense of humor to call them _silly_. I, and many others, see these as outright criminal, dangerous and outrageous. Hopefully, Europe will steer clear of such atrocities (ditto for software, albeit those are not criminal, nor *that* dangerous, but outrageous still are). Hopefully...

    Can you guess I leave in Europe? :)

    ReplyDelete
  6. "Or even worse, comes after you with an infringement suit because they patented the idea."


    Impossible because of prior art

    ReplyDelete
  7. http://www.infoq.com/news/RedHat-Sued-Due-to-Hibernate-3-O


    I wonder if ECO is next?

    ReplyDelete

Please keep your comments related to the post on which you are commenting. No spam, personal attacks, or general nastiness. I will be watching and will delete comments I find irrelevant, offensive and unnecessary.