Charles Arthur: The patenting of software is a complete mess and discourages innovation

So, congratulations to Tim Berners-Lee - now Sir Tim, after his achievement in inventing the World Wide Web all those years ago in 1991, while tinkering on a NeXT computer in search of a way for all the scientists at CERN, the European particle research lab, to collaborate on each others' work.

You can find a regenerated version of his original proposal online (at www.w3.org/History/ 1989/proposal.html); if from that you can envision a globe-spanning, cross-platform, client-independent, life-changing system, then you're better at interpreting diagrams with lots of vague lines on them than I am.

Arguably, though, what made the Web spread so fast wasn't just that it let you click about from place to place; that had always been possible using protocols such as "gopher" and plain old file transfer ("ftp"). The key thing was that there was no restriction on using the idea embodied in the software that runs the Web. Tim Berners-Lee never tried to patent the idea of the Web.

Software patents are, in fact, a live issue at the moment, though mostly invisible to the vast majority of people. The problem is that there's a lot of good sense talked on both sides, both pro- and anti- the patenting of processes done by computer.

Here's the argument in favour. We have had patents for more than a century; the idea is that you produce an "invention", and in the patent application describe how it works. In return, you get legal protection for a certain length of time to discourage others from copying it (if they do, you can take them to court). When the patent expires, anyone can use the idea. Thomas Edison patented hundreds of things, including the light bulb and the wax cylinder system. More recently, the pharmaceuticals giant Pfizer has benefited from its patent on Viagra: rivals knew the formula but couldn't, on pain of being sued, copy it.

Since then, we have moved a lot more of our business processes on to computers. If you did any shopping at Christmas, some of it was bound to be online. Those shopping sites, and all the other computerised business processes that you don't see but benefit from, are the light bulbs and wax cylinders of the modern age, so surely they should receive the same protection as blue pill makers? In the US, the Patent and Trademark Office has agreed since the Nineties, and in 1999 awarded Amazon a patent for its "One-Click" shopping system, whereby you can purchase something on its website by hardly lifting a finger. It wasn't the only one - that year the USPTO awarded 1,016 internet and network-related patents. (One-Click hasn't been patented as such in Europe; in 2001 it was split into two patent applications, of which the European Patent Office has granted one, and left the other pending.)

Here's the anti-case: who's most likely to take out patents? The big organisations, which can afford the research and the patent lawyers to draw them up. They can use these to attack upstart rivals, claiming that they infringe their patents, and sue them to death. Yet it's the upstart rivals that tend to turn the software world upside-down: think Netscape. Think even Microsoft, which was small once too, when it provided the operating system for IBM's first PC.

And patenting software - pure software - suffers from the problem of abstraction. What exactly do you leave out of the patent? Your patented implementation needs to run on a computer. What computer? Running what software? Does it rely on the software to work? In which case, must you also be able to patent the software that it needs, too?

One starts to wonder whether, if Amazon had built One-Click (which it applied for as "A Method and System for Placing a Purchase Order Via a Communications Network") on top of Microsoft software, it would need to specify that Microsoft product in the patent. If so, rivals could implement One-Click without infringing the patent by using different software. In other words, patenting software is a complete mess; what's more, it discourages innovation - being able to copy an idea and use it again yourself - the purest form of intellectual bootstrapping - will make any useful technology spread at dramatic speed. And surely we want everyone to benefit?

Even one of Amazon's founding programmers, Paul Barton-Davis, said that "Amazon.com's early development relied on the use of tools that could not have been developed if other companies and individuals had taken the same approach to technological innovation that the company is now following."

The European Patent Office is leaning towards an "in-between" system. Being patentable means "making a technical contribution" - that is, "a contribution to the state of the art in a technical field which is not obvious to a person skilled in the art".

Quite where you find this programming equivalent of the man on the Clapham omnibus (the judicial phrase for the average person) is anyone's guess, which is part of why open-source advocates dislike the idea.

Sir Tim has already made his dislike of software patenting well-known. In May 1999, he said that patents "are getting in the way of a common, universal Web". He returned to the fray with zest last October, when a company called Eolas - which claimed to have a patent on the idea of embedding content (such as video or audio) in a single webpage. (It filed in 1994, and received it in 1998.)

By suing Microsoft for making Internet Explorer able to play such content, Eolas went after the biggest possible fish. Microsoft vowed to change Explorer - but that will make it incompatible with those old pages. And Explorer is the most-used browser on the Web. That means huge changes for millions or even billions of web pages to be viewable. But Sir Tim says the idea of embedding content was put forward in minuted meetings between other people in 1993. That would invalidate the patent.

The USPTO is expected to pronounce on the case sometime in the next month or so. One has to hope that it will see sense, and go with Sir Tim, and annul the Eolas patent. I'd rather see his path followed: he passed up the chance to be a multi-millionaire, but we are all enriched just a little bit by his invention. By contrast, could you name a single person at Eolas?

network@independent.co.uk