Monday, January 11, 2010

The patent process is fundamentally flawed

A patent is a way of telling the world that you had a great idea first. They work just fine in most cases but the big problem with ideas is that the patent grant must be made by an examiner who is expert in the field of the problem domain. And, as patents are often new ideas, there may be no experts other than the person who defines the patent itself.

Patents have to be non-obvious. We cannot patent the process of breathing for example, although I'm sure a healthcare company in the US is probably working on that right now. This leads me to the i4i patent which describes a method for storing word processing documents in an XML (or, if you read the patent, SGML) format.

Today, Microsoft have been forced to stop selling Word that saves in docx format because it supposedly violates the patent of i4i's document storage method but in reality, the patent should have never been granted in the first place because it does nothing more than describe a possible algorithm for storing data in an open and freely usable format. XML by definition is useful for storing absolutely any data so patents based on specific uses of the XML format are obvious and so should never have been granted in the first place.

This ruling, like so many rulings or grants on algorithmic processes, have obviously been made by persons with no domain knowledge and are completely arbitrary.

I believe that patents should be restricted to physical and tangible objects or manufacturing processes and that patents on algorithms and data codes, especially genetic ones, should be banned, relying instead on copyright law for the former and ownership of the original material for the latter.

No comments: