If the term "bullish" ever defined a character, it would be Steve Ballmer. Loud, energetic, almost evangelical, the former Microsoft numbers man and now chief executive officer harnessed all his strength last week at Microsoft's Redmond, Washington headquarters to deliver the company's new MSN.net strategy to the world's press.

His booming voice crescendoed to a squeaky pitch at times as he excitedly told the gallery that MSN.net was the result of a full-time effort on the part of Microsoft since Bill Gates stepped down as chief executive officer and became chief software architect to establish exactly what the implications of the Internet were and where the opportunities lie.

He said the company had been like a "caged beast" in recent weeks wanting to "burst out of the closet" with its new strategy. Microsoft's PR advisers had insisted on deferring the event from early June when it became apparent it would coincide with the final court order to split the company in two.

Now the future had been unleashed and, with his boundless enthusiasm, Mr Ballmer was leading the charge. Excitedly he outlined how Microsoft was going to transform itself from a software to a services business. He urged patience among Microsoft customers as "this stuff isn't going to happen in the next day or two", but the result would be a simplified user experience "supporting the next generation of the Internet as a platform, a user interface and a set of services".

The initial reaction among many of the sceptics gathered was: "So what? Isn't every company out there reinventing itself around the Internet?"

Every so often you have to remind yourself whether you like it or not that Microsoft is not just A.N. Other software company. Its software provides the basic operating platform for 90 per cent of the world's computers and its applications are run on a large proportion of these machines.

When it announces it is going to introduce a "dot.net" version of Windows, Office and the MSN websites, it's going to make a difference for a lot of people who use these products.

For example, Office.net will allow subscribers to rent Word, Excel and PowerPoint and use it via a Web browser on their mobile phone, hand-held device or car computer.

This is Microsoft's answer to the dual challenges of the rise of Internet-based services and the proliferation of new computing devices. Microsoft.net promises to be a hybrid service, which combines the speed of remotely located powerful servers with the local power and storage capabilities of appliances like wireless phones and hand-held digital assistants.

So what's the difference between this and the "thin client" strategies of Sun Microsystems or Oracle?

Asked later if this was not just a case of Microsoft coming late to the Net party when Java is well established, Mr Ballmer looks like his head is about to rotate off his neck. He says: "Java is nothing. Nothing. It has nothing to do with the user interface. There is no Java work being done in XML any more. It's only a programming tool for server applications, where I admit it's a decent competitor - but it has completely disappeared from the client or user end."

True, even if the success of Microsoft.net cannot be measured for another couple of years, the company has made a concerted effort to extend the Internet platform to all users of computing services.

The combination of server hosting with smart applications for mobile devices is, in theory, an all-embracing strategy. Similarly, the company's adoption of the increasingly popular XML (extensible mark-up language) computer language poses a direct challenge to Sun Microsystem's Java product, even though Microsoft is too proud to admit as much.

With so many small innovative companies out there already using XML to develop software, can we expect a slew of acquisitions by Microsoft? Mr Ballmer smiles wryly and says: "Yes you can expect a lot of acquisitions but if I am any more specific, it will end up costing the company more."

Speaking of costly ventures, what's Mr Ballmer's reaction to the order to split the company in two on the basis of anti-competitive practices. Without pause for thought he responds: "I'm sure there's a high-minded legal aspect to all of this, even if it's misshapen and misdirected. There's another response that could say these are the ways careers get made in legal circles."

Rumours of the death of Microsoft's arrogance are greatly exaggerated. As long as Mr Ballmer is at the helm of this giant, it's going to go in fighting every time.

There isn't even a trace of irony in the fact the MSN.net strategy is a mandate to dominate the market for Internet-based applications. Doesn't this all sound a bit familiar?

One analyst recognised as much amid all the hoopla in Redmond last week. "This is Microsoft full speed ahead, damn the torpedoes. But if the justice department doesn't look at this and say it proves their point, I'd be surprised," said Mr Chris LeTocq, a Gartner Group analyst.

As it stands, a number of shortterm restrictions on Microsoft's business practices have been suspended pending the company's appeal of the court ordered break-up. However, if upheld they could severely hamper the company's progress with Microsoft.net.

Mr Ballmer says, one way or another, the company plans to finish work on Microsoft.net. "It would be better for consumers if this product was tightly integrated, but we will definitely finish it."

There are two critical obstacles to MSN.net repeating the success of Windows.

Firstly, will be the willingness of software developers to unite behind the new platform and write applications for it. Without that MSN.net will be a pipedream and, in its time, Microsoft has managed to disenchant its fair share of developers.

Secondly, commercial applications of Microsoft.net products are at least two years away. Meanwhile, the underlying integration of the Internet and software on its MSN.net platform does not bode well if Microsoft loses its appeal and has to split into two distinct companies.