Devil's Advocate: Microsoft's foolish patent policy

Microsoft's XML-related patents seek to convince the public it believes in open formats while requiring them to pay for access to their data. The solution, says Martin Brampton, lies with the user.

Is it really true that Microsoft has patented the transformation of objects into XML? That would certainly be proof, if proof were needed, that so-called intellectual property has little to do with innovation.

Software objects involve ingenious ideas that have had growing attraction. The concepts were always sound, and they were all invented before Microsoft even existed. But their use was limited because applications were restricted in scope and computer power was scarce. As those factors changed, so objects became far more central to computing.

Likewise, although XML itself is a relatively recent phenomenon, it is based on SGML. And SGML was also invented long before the foundation of Microsoft. Again, it was a question of timing, as the tantalising goal of building self-describing data looked to be achievable with rising hardware capabilities.

It is reported that Microsoft has now achieved a patent on the transformation of objects to XML and vice versa. This is a pretty bizarre thing to happen. As an industry analyst, I became aware of XML as soon as it started to make waves. My immediate reaction was that XML documents were objects without the behaviour.

Now that was hardly an astonishing insight. I would never have expected to rush off and patent it, even if I could afford to patent anything. It was an insight that was common to anyone who had at least half an understanding of both objects and XML. I must have discussed the point with numerous people, and most likely spoken about it at public conferences. If it is a patentable idea, then the whole patents system is demonstrably absurd.

If large companies are to be granted patents on ideas that have been commonplace for years and are based on fundamental concepts well understood for decades, they might as well be given taxation powers. The idea that their revenues are hard won in competitive markets will be defunct, if it is not already.

The reality is probably just a little more complex. The patent is most likely less wide-ranging than recent reports would suggest. Microsoft's efforts in relation to XML patents have mostly been concentrated on achieving two contradictory goals. One is to give the impression that Microsoft software, especially Office, is storing its data in an open format. The other is to make it impossible for other software vendors to handle data from Microsoft products, except perhaps after making substantial payments to Microsoft.

Provided Microsoft restricts itself to such relatively limited goals, it may well succeed in exploiting the patent regime, which is anyway slanted towards the interests of large companies. The only remedy lies in the hands of the pusillanimous users of packaged software.

It is extraordinary that so many organisations have permitted so much of their information to be locked into proprietary mechanisms, usually without a second thought. It is a commonplace that information is a valuable asset, yet people seem carelessly to give others the right to control it.

A reasonable starting point for a balanced commercial negotiation for software would seem to be that the data belongs to the organisation that creates it, not to the software supplier. The creator should be prepared to pay a reasonable price for the use of software. However, a prudent software purchaser would insist that the data will always remain fully accessible, even if use of the software ceases. Simply accessing one's own data should not be grounds for further payments.

From a mixture of carelessness and inequality of negotiation power, software buyers have generally failed to implement this principle. In the end, they pay a high price for this. Just how high the price can go before provoking retaliatory action remains to be seen.