• silicon.com
• Technology
• Software

By Martin Brampton, 18 September 2007 15:56

COMMENT

Microsoft is a pioneer in pushing anti-competition laws to their limits, argues Martin Brampton. The real lesson of this week's EU decision is that it is pointless issuing a court ruling 10 years too late.

Commentators are saying Microsoft has suffered a stunning defeat at the hands of the European Union's second highest court.

This is a gross overstatement. What has happened is no more than a feature of the standard business approach of Microsoft and many others.

It has long been held that free markets fail when monopoly power exists. Legislation to promote competition goes back more than a century.

Despite Microsoft's assertion that US antitrust legislation is obsolete, the company was itself a beneficiary of earlier action against IBM. At the time, IBM was selling 70 per cent of all computers worldwide, and operating software was bundled with the hardware.

IBM was then a massive bureaucracy which went to great pains to comply with onerous demands imposed on it for record retention. The company became increasingly sclerotic until John Akers, the last of the old guard, was forced out as a result of huge losses.

Lou Gerstner introduced a new era but by then the world had changed. Even at its most dominant, IBM does not seem to have pushed boundaries to the extent Microsoft has done.

This is graphically illustrated by comparing the circumstances now and then. IBM's 70 per cent share of the world computer market compares with recent comments from European competition commissioner Neelie Kroes: "A market level of much less than 95 per cent would be a way of measuring success."

And although bundling was outlawed as part of the IBM judgment, the fact is that most PCs are sold with a copy of Windows included.

Strictly speaking, it is possible to decline to accept the bundled copy of Windows that comes with almost every mass-market PC. A few people have actually done it.

It takes a lot of very careful marshalling of evidence and a struggle to convince sceptical PC vendors that you are entitled to a refund.

With a large PC maker like Dell, after several hours of painstaking work and a certain amount of frustration, you may receive something like £25 back. Most people don't bother.

So it seems that competition regulators start from a position that is far more extreme than the IBM situation decades ago. It will take more than a single court decision to alter the trend.

Microsoft has had a near monopoly for at least a decade, and the EU court case has been running for years. It would be extraordinary if Microsoft were not fully prepared to lose this case and to have plans for moving on.

The real question is whether the EU has the ability and determination to follow up with considerably more action to promote competition. If it does, the result is likely to be conflict with the US because of two pronounced tendencies.

One is the reluctance of the US authorities to take effective monopoly-busting action against its own companies. The other is the extremely restrictive character of US policy on so-called intellectual property rights.

Currently, software is primarily protected by copyright law. US legislation permits software licences to bar decompilation, a restriction not permitted in Europe.

It does seem odd that copyright can be claimed in something that is kept secret but there we are. Reverse engineering is often considered a vital tool for legitimate competitive activity related to follow-on innovation and interoperability.

Worse are the potential effects on software of the US attitude to patents, which has opponents even within the US. Currently, the restrictions on what is patentable in the US are extremely weak. A patent can be awarded to anything that is shown to be non-obvious, to include an inventive step and to have industrial utility.

In reality, even these loose guidelines are interpreted to permit surprising patents. Priceline.com was notoriously able to patent the idea of allowing consumers to propose transaction prices, despite this being a practice that dates back thousands of years.

Given such a loose definition of what is patentable, it is inevitable that much computer software qualifies. Implementing perfectly straightforward systems often throws up quite complex problems. Software developers frequently spend hours working out the code for a neat solution.

The fact that it does take time and effort, even for a skilled developer, is clearly proof that the solution is non-obvious. Once that has been established, it is a small step to demonstrating inventiveness, especially if much of history is disregarded.

And it is in the nature of commercial software development that the results have industrial utility, at least in a general sense.

It therefore seems that the average software developer will come up with US patentable ideas quite regularly. In most cases, they won't have the resources to patent them. A regime of this kind obviously favours large companies that have the capacity to handle numerous patent applications.

Since US patents are based on first to register rather than first to invent, the poor developer who has no patent attorneys at their disposal is doubly penalised. Innovative work is not patented, and the developer may in the end be barred from using their own innovations.

Microsoft has recently been brandishing its claim to own patents on Linux software. That the company owns patents breached by open source software is perfectly possible. It is possible even if the innovative features were first created by open source developers and not by Microsoft.

That is the nature of current US patent practice, which many now see as going beyond what is justified. After all, intellectual property rights cannot be based on some kind of natural justice, they have to be based on what will give the best overall advantage to society as a whole.

Predicting Microsoft's next move is pure speculation. But exploitation of patent law looks to be one direction the company will regard as showing considerable promise.

A single court defeat is only the price companies expect to pay when they push to the limits and beyond. And as the defeats are to be expected, so too are plans for developments beyond them.

The justification for EU actions against Microsoft and others are that excessive concentrations of commercial power stifle innovation and lead to excessive prices.

But unless the EU is to find ways to step up its activities and to bring each case to a conclusion far more quickly, it seems unlikely that its victories will be anything more than transient blips on the progress of giant corporations.