Software patents - time for a new Book of Bounty?

Patents are old. The first known English patent was granted in 1469 by Henry VI.

Their purpose was to grant monopolies for the manufacture and sale of commodities. In 1610, King James I declared in his Book of Bounty that 'monopolies are things contrary to our laws'.

Section six of the Statute of Monopolies (1624) rendered illegal all monopolies except those 'for the term of 14 years or under of new manufactures ... to the true and first inventor', provided that they were not 'contrary to the law nor mischievous to the State by raising prices of commodities at home or hurt of trade'. Patent No. 1 was granted in 1617.

Patents now provide protection for 20 years.

The idea that they can be applied to software would be laughable but for the sad fact that they are here now. With large numbers of patents at their disposal big companies are able stifle the innovation of smaller companies by requiring them to do ever more to ensure an idea hasn't already been patented.

The sheer weight of annual applications shows that companies are desperate for patents to augment their commercial armoury and avoid having to pay licence fees to someone else - the quid pro quo defence.

Large software companies have much to answer for. Instead of acting responsibly they are continually filing patent applications of the most speculative nature at various offices worldwide in the hope that some of them will stick. Microsoft alone sent in over 3,000 in 2004.

Common sense has no place in the current system. Like many I do not believe that it can ever be right to grant a patent for an algorithm - it would be like patenting a particular way of driving a car.

Yet the GIF patent was granted and patents for a/v codecs abound - indeed the BBC developed its own codec (1).

Another consideration is the practicality of granting patents. With the avalanche of applications that are arriving on their desks, patent clerks cannot be expected to act with due diligence and discover if a claim is valid.

There are tens of thousands of software developers writing code all over the world. The idea that these developers are simultaneously writing code that is new, original and not patent-infringing is risible.

As a software developer I want to be able to write code that performs the task required by the specification.

I am quite capable of developing new algorithms and resent the fact that because someone else has already patented a particular algorithm I am not then allowed to use my independently invented version of it.

In practice, developers (certainly in small to medium-sized companies) ignore patents altogether. If they didn't the development timescales would be too long and the cost to the company too large. The situation for open source developers is even worse (2).

Some companies, IBM for example, have stated that they won't use their patents against open source software, which is gratifying to hear.

However a system that depends on the benign actions of big companies to prevent new competition being strangled at birth has no place in a modern competitive world. Software patents should be scrapped and all existing software patents nullified - we need a new Book of Bounty!

Simon Geard MBCS CITP

Would software patents be all bad?

I once ran a software company, making intelligent internet software, and we employed a lot of software professionals. We needed to pay them, every month, so that they could fund their families, their mortgages and live properly.

We needed to invest in these people, and in our technologies, and in our products, to develop them and retain the flow of money that paid their salaries.

But if we couldn't protect our work, couldn't benefit from our insights and inventiveness, it would be hard to justify all this investment.

This seems reasonable, especially in a fast-moving field in which re-engineering solutions that others have created is not that complex.

Patenting can provide that protection; protection can foster growth; growth can encourage innovation.

Without protection there is potentially no benefit in investing in development, and hence innovation could wither and die.

Whether the current software patenting proposals, and the institutions to oversee and run the process, are up to the job is a different matter. But there are arguments that patenting is not all bad.

Russell Beale MBCS

Further reading

Two sites reflecting the opposing views are:
http://swpat.ffii.org (against)
www.patents4innovation.org/index.php/eng/patents_cii/ (for)

References

1. BBC
2. No Software Patents

This article first appeared in May 2005 ITNOW.