Mark Blafkin is confused about my analogy between eminent domain and software patents, which probably means I didn’t explain the analogy very clearly. So let me see if I can be more explicit.
The way modern “redevelopment” projects work is that a large developer will go to a city and say “We would like to develop a new shopping mall/office park/apartment complex/whatever in such-and-such a neighborhood. But we’re only willing to do so if you give us control over all development within that neighborhood. We don’t want to worry about some other company building something in the neighborhood that we didn’t plan for.”
The city will then scrutinize the application, go through some legal technicalities such as declaring the neighborhood to be “blighted” (which, with enough ingenuity almost any neighborhood can be), and then sign a contract with the developer that essentially gives the developer a monopoly on development in the area. Any property owners who refuse to go along with the developer’s plans are removed using eminent domain.
Now, a company like Verizon will go to the patent office and say, in effect, “We would like to develop a new VoIP application. But we’re only willing to do so if you give us control over all development of VoIP applications like ours. We don’t want to worry about some other company building a competing product that we didn’t plan for.”
The USPTO will then scrutinize the application, go through some legal technicalities such as deciding whether the patent is “obvious” (which, with enough ingenuity almost any concept can be made unobvious), and then issue a patent to Verizon that essentially gives the them a monopoly on the development of VoIP applications. Any software companies that try to build a competing VoIP application are driven into bankruptcy with a patent infringement lawsuit.
Mark, there are some similarities here, don’t you think?
Frankly, I think Mark’s retort proves my point:
However, he should have to consult the owner of the land before he starts building on it.
We know that if a real estate developer wants to figure out who the owner of a piece of land is, he goes down to the county records office and looks up the entry for the land in question. While there are a few legal technicalities, a real estate developer can usually find the person—and there usually is just one person—he needs to talk to to get permission.
What’s the equivalent step for a software developer?
I’m seriously interested in what Mark has to say about this. Because it has me stumped. There certainly isn’t a single “property owner” you can talk to to get rights to all of the patents that might be implicated by a given software application. Moreover, there isn’t even a reliable procedure for getting a list of all of the dozens of people who hold patents that might be relevant. The closest thing is to spend tens of thousands of dollars hiring a patent lawyer to do an exhaustive patent search. But that will just give you a list of dozens of patents that might implicate your own. Many of the patents will be vague, only tangentially related to your software, or so broad that courts are likely to invalidate them. If you tried to negotiate licenses with all of them before getting started, you’d never get around to developing your application.
Ordinary property rights reduce uncertainty by creating clear rules about who you have to negotiate with before beginning a real estate development project. Patents, in contrast, cast a cloud of uncertainty over a software product by giving dozens of different people the potential to sue you for patent infringement with no easy way to find them and no way to be sure if a given patent covers your product or not. It’s downright silly to equate the two.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.