Ronald J. Mann has published a paper with that title on SSRN (link credit: Axel H. Horns).
I happen to believe that there are so many software patents around that nobody actually trying to observe them all can get any programming work done, which is about what Shapiro teaches us in his paper about "Navigating the Patent Thicket".
So I was interested to have a look at why I might be wrong.
Mann's argument:
He has talked to a lot of people actually involved in the use of software patents in America. One of his results is that none of the start-up firms he has spoken to do any prior-art searches before beginning development of their products. (Page 53).
I have no problem whatsoever agreeing with Mann here.
Of course nobody does any prior-art searches. There are two reasons for that. Patent information is utterly useless for people who actually want to develop software, and is deliberately kept useless by those drafting the patents, for example by filing software listing as a paper copy. And, more importantly, actually researching prior art leads to bad faith if one finds anything, and stronger sanctions in the infringement lawsuit later on.
However, if the problem of software patents could easily be solved just by ignoring them, nobody would worry. Just shutting your eyes and pretending not to see the patent thicket does not remove its existence.
Mann then proceeds to explain that infringing software patents is no big deal because IBM is very unlikely to actually sue infringers (Page 53).
Again, this is only a denial of reality.
Of course, if every software patent holder started enforcing every patent, the whole software industry would collapse just in the same way as it would if everybody actually started researching prior-art and asking for permission all around the place to actually get some development work done.
But, while it is nice to see that IBM doesn't enforce all its patents all if the time, that changes nothing about the fact that they could do so whenever they happen to feel like it.
The fact that no nation has actually used a thermonuclear weapon of mass destruction for the last fifty years does not mean that those weapons are a myth. In exactly the same way, the fact that not all software patents are enforced does not mean that they are a myth.
After spending several pages on explaining why IBM won't enforce software patents, Mann then tells us that IBM or Microsoft are going to enforce their software patents after all, by demanding royalties from just about any software project, since anyone can expect to be violating some of the big player's patents (Pages 57 to 58).
That obviously is in contradiction to what he just told us a few pages back. Either IBM won't enforce any software patents or they will ask for royalties. It would be nice if the author could make up his mind.
Mann tells us then that even if IBM can ask any other software developer to pay royalties, that still doesn't qualify as a "patent thicket", since developers are still able to secure the necessary licenses (Page 58). To quote:
On the contrary, a patent thicket would exist only if industry licensing practices were such that firms in the industry commonly were unable to agree on terms for licenses and thus retreated from the field of innovation."
Obviously, if you can get licenses, you don't need to worry about violating patents. And the Shapiro paper quoted above describes cross-licenses and patent pools as strategies for dealing with patent thickets.
Again, the fact that you can navigate the patent thicket by getting licenses doesn't negate its existence.
As a result of looking at Mann's paper, I don't find anything not compatible with my belief that the large number of software patents granted already would shut down the software industry completely if all patents were enforced at all time (with the exception, of course, of large companies able to pay $440 million in settlement fees for patent violations).
Posted by Karl-Friedrich Lenz at April 14, 2004 02:49 PM | TrackBack