You have to feel a little sorry for Microsoft.

Many of you will now be preparing an email to point out that no company deserves sympathy when it is sitting on nearly $60bn in cash and has been accused of abusing its market dominance in the US and Europe.

But the fact that the software giant has few friends is not the reason for the fundamental problems it faces over the next few years.

Redmond faces a huge cultural shift, not unlike the changes IBM had to go through in the early 1990s after its traditional mainframe market was undermined by the growth of personal computing.

The first cracks are already showing. Longhorn, the next version of Windows, was touted by Microsoft executives as the most important release since Windows 95, a new architecture for the operating system.

But last week it was announced that one of the most important new components, the WinFS file system, will no longer be part of the product. Its development has been delayed by the extra effort put into the extra security features in the recent XP Service Pack 2, and will come as a bolt-on at a later date.

Of course, critics will say that Microsoft is simply doing what it should always have done and making product security a fundamental part of its software.

But the company's success has always been founded on a culture of 'gee whiz, that's great, build that into Windows.' The focus of development has always been on new features, more features and, hey, let's put some extra features in. Its eager developers have been given free reign to make Windows as full and functional as possible. Quality coding was a factor, but not a priority. It was assumed that PC users would put up with a few bugs in return for all the great features and a competitive price.

But the internet came along and changed everything. Suddenly, millions of Windows-based PCs are exposed to the world, open to every miscreant, hacker and envious techie.

Now, Microsoft has to completely change the way it approaches software development. It has to tell its developers to do boring things such as test, test, and test again. Then do some more testing, and finally, test one last time.

Don't think this is a problem for Microsoft alone - every software supplier in the world has traditionally worked on the basis that software is never bug free. But not every software supplier in the world is exposed in the way Microsoft is. Be sure that most of them are quietly thinking - there but for the grace of God go I.

Do not doubt that Microsoft is taking security extremely seriously - the effect on a product as vital as Longhorn proves that. But the company is fighting its own legacy - an enormous, monolithic operating system that has grown organically for years, and frankly would never be developed in the same way if it was created today. If you wanted to get where Microsoft wants to go, you wouldn't start where it is now.

PCs are consumer products. Who would buy a HiFi that needed a monthly download of additional software? Redmond is trying to find ways of easing the bug fix burden - the automatic update function in XP is an example - but it doesn't reflect the way ordinary consumers use PCs. A friend recently bought a brand-new XP-based PC. The first time he switched it on and connected to the internet, the auto update feature said there were 32Mb of patches to be downloaded. Home PC users simply won't bother, so their PCs are left unprotected, they get hit by hackers, and everybody blames Microsoft.

Changing a software product is one thing, changing the mindset of the people doing the development is a whole other challenge. It took IBM 10 years, and was only completed thanks to a hard-headed outsider, former chief executive Lou Gerstner, who was willing to question the status quo and not afraid to make controversial changes. Does Microsoft have anybody who can do the same? And as the company comes under new competition in areas such as mobile phones, broadband content, digital TV, handheld computers and home entertai nment, can it juggle the cultural change with the pace of technology change? It can, and it will, that $60bn will help. But the effect on future product development plans will be significant, and could fundamentally change Microsoft's relationship with IT directors.

What do you think? Email feedback@computing.co.uk

If you want to be first with the news, visit Computing every day.