Today is No Software Patents Patents day. I go both ways on software patents. As a business owner I like the idea of asserting an “unfair advantage” over my competitors by coming up with some super-duper patented idea that completely locks in my market. As a software engineer I grow concerned at the land grab tactics of some companies. As a member of the world community I grow concerned at WIPO’s and latterly GATT’s attempts to enforce patent law globally over countries who are coming late to the patenting party.
So first a little history. Much of what follows is based on how English patent law developed.
Patents developed from monopolies granted by the King of England in the 1500’s to businesses such as the manufacture of commodities such as soap, leather, salt, glass and knives. Needless to say this granting of monopolies was abused by subsequent kings. This led to to revocation of patents in 1610. This was followed in 1624 by a statute of public interest that banned all monopolies barring those related to “…the sole working or making of any manner of new manufactures within this Realm to the true and first inventor“. The period of monopoly was 14 years. This was probably the first legal patent act.
The 18th century saw the addition of a requirement that a patent application include a specification of the invention. Two other clauses were added that would become relevant later, the possibility of improving an existing patent and the possibility of patenting an idea provided it those ideas “were clothed in a practical application”.
The 19th century saw much streamlining of a complex and inefficient process, resulting in the creation of a formal Patent Office and a basic patent examination scheme. In the 20th century we added novelty to the bar for entry and codified existing patents to allow some kind of searching.
Why have Patents?
But what are patents good for? Well the historical goals were to encourage innovation and invention and to promote disclosure of that invention. More disclosure leads to more rapid development of society as one invention contributes to the next. Without the patent system disclosure often led to inventor losing the ability to exploit his invention for gain, as others copied the invention without payment or license. So patents grant a monopoly to the inventor in order to prevent others from copying his invention without his/her permission.
This worked reasonably well while the inventing process was laborious (99% perspiration, 1% inspiration), time consuming and expensive. It turns out to be less useful when patents can spring from peoples’ heads in a matter of moments, this is the case with software patents.
Why are software patents problematic?
The first software patent was granted in 1962 for a memory management application. Since then there has been a explosion of software patents and business process patents. These ideas spring to life practically instantaneously or are mined out of a body of work after the fact.
Both these kinds of patents differ from those that proceeded them in that both the tests for novelty and prior art are impossibly difficult for a patent office to ascertain. In most cases this is limited to a search of the existing patent repository and/or a cursory search of some of the available literature. The problem arises with novelty due to the breadth of the field of software and the fact that the industry is changing so fast.
The problem with prior art arises because discovery of “art” is not enough, there then has to be proof that the “art” predates the current patent filing. This is actually incredibly difficult to do for even a small amount of material.
The “patent industry” which consists of the solicitors and lawyers who file patents and the national organisations that grant patents conspire to increase the problem. Both these groups of people are motivated to create more patents. The bottom line being if you give a patent lawyer 20,000 dollars you are going to get some kind of patent through fair means or foul.
All these patents are buried under piles of “claims” written in legal speak which utterly defy the best efforts of any kind of search. I challenge any person to produce all the patents related to spreadsheets that have been filed since Visicalc was created. (Note that Visicalc contains no patents, but what if it had?).
So we have patents that are poorly approved due to difficulties with novelty and prior art, an industry that depends on creating more patents and a existing repository of patents that are not searchable.
The Real Problem
All this pales into insignificance behind the real problem with patents. Patents were designed to grant specific rights to the to “the true and first inventor“. The granting of a monopoly (which a patent is) should never been done lightly in a free trade economy. The framers of original patent law believed the goals were to grant the monopoly on the basis that society as a whole benefits from the disclosure of inventions.
The real problem is we have turned patents on their heads and made them fungible, tradable commodities that can be bought and sold on the open market. To the highest bidder goes the largest pile of monopolies. Nathan Myhrvold has built a global business on this basis and Marshall Phelps has spent years at IBM and Microsoft turning their patent portfolio into cash. This aggregation of monopolies cannot be what the framers of patent law intended.
So disclosure of software inventions is not served by the patenting process (when was the last time you searched the patent archive for a solution to a programming problem). The original inventor is not served (the vast majority of software patents are filed and assigned to corporations) as he rarely retains the rights to his invention. Ssociety is not served because society has a whole derives no benefit from the patenting process.
If none of these parties are served what value are software patents?
What can we do?
The patenting system is here to stay and patenting has a real value in areas such as bringing new drugs to market where the costs are so astronomical and the failure rate so high that some method of recouping the costs must be put in place to encourage private sector involvement.
So some recommendations.
- Patents must result in execution of an idea: A patent must have a use it or loose it clause that drives an inventor to either develop the invention or license the invention. Patents not used in such a way expire in much shorter time period. A patent filing must be followed within three to five years with a description of a business articulation of the idea or the patent expires automatically
- Patents expire faster when they are transferred: Patents expiration date should be shortened (by 1 year? 5 years?) by each change of control. This will discourage patent trading and encourage the inventor or someone he sells the patent to at an early stage to exploit the idea.
- You can’t sue an infringer unless you are the inventor or executor: Only the original inventor (if he holds the patent) or the executor of the real idea can sue a potential patent infringer.
- The Patent Office manages the process: We remove the lawyers from the process by shifting much of the burden of drafting and approval to the patent office appeals and infringement proceedings are also managed by the office rather than the courts. The whole process is funded by the fees required to file. So the 20k we used to give to lawyers goes direct to the patent office.