Is Microsoft's .net a trap?

There is bound to be cynicism about Microsoft's .net strategy. Microsoft promised "an advanced new generation of software that melds computing and...

There is bound to be cynicism about Microsoft's .net strategy. Microsoft promised "an advanced new generation of software that melds computing and communications in a revolutionary new way". What it delivered last week was, essentially, XML functionality for three key product upgrades: Windows 2000 Datacenter, Exchange 2000 and SQL Server 2000.

When a supplier announces a quantum leap that will shape the next decade of IT, but briefs its own people that 25% of the code is already written, users are right to switch on their hype-detectors.

But the "vision" part of the .net strategy is important. It is Microsoft's survival strategy, prompting the question: if it works, how will it change computing?

At the heart of the .net strategy is the insistence that Microsoft will keep its proprietary talons off the emerging open data standards of the Internet economy. Microsoft's surrender in the long-running meta data standards war last week was a token of good intent. So is its repeated reassurance that .net is built on XML, which is policed by the World-Wide Web Consortium (W3C).

If .net is delivered, we will see the transformation of development tools, operating systems, and end-user devices at the high-volume, low-cost end of business computing: it will be a distributed computing environment, with multiple I/O devices and interchangeable components that ensure common standards across proprietary software and hardware platforms.

The problem is that Microsoft's path to that nirvana lies through extending the XML functionality of its products and services in a way that threatens open standards.

We have seen what can happen with simple browsers: HTML is also policed by W3C - but that did not stop Microsoft "embracing and extending" the language beyond its agreed standard format, with successive versions of Internet Explorer, in a bid to wipe out rival browsers.

With XML, the opportunities to gain a proprietary stranglehold over an ostensibly open standard are even greater. HTML is a series of pre-defined tags whose basics you can learn in an afternoon: XML is a blank sheet, currently in the process of being filled with sector-specific meta data in a variety of global standards conferences. XML would always be more complex and more difficult to police, even if there were no large IT corporations circling round it like sharks.

Microsoft thinks the biggest thing wrong with Internet MkI is that it is "read-only". You can read the Web, but not rewrite it in real time. And Microsoft slams today's browser software as "a glorified read-only dumb terminal".

But that's what makes it the Internet: the ability to access information through an open standard protocol, using open standard client software. Sure, it's "read-only", but it is also "write- easily". The static state of the Internet is what allowed it to take off as a revolutionary one-to-many form of mass communication.

The Internet MkII should be built on open standards, policed by user coalitions not IT suppliers. The alternative would be a series of sophisticated point-to-point closed networks based on IP. There is a lingering suspicion that this is where .net will lead.

Remember: if you kill the open standards that have allowed HTML to conquer the planet, you kill the goose that lays the golden egg.

Read more on IT architecture