In 2001, Michael Robertson sold mp3.com to Universal for $350m. This year, the flamboyant executive extracted $20m from Microsoft to settle a court case Redmond had instigated over Robertson's decision to name his Linux start-up's software 'Lindows'.

Now Robertson is trying to inspire PC makers to pre-install his company's desktop Linux software, renamed 'Linspire'.

I can buy Linux for the desktop from established parties like Sun Microsystems, Red Hat or SuSE. Why should I go to a start-up company called Linspire?

There aren't any companies out there like ours. When we talk about companies that are making real investments in desktop Linux it's few and far in between.

You have the ones that have announced it, but don't do anything: Red Hat, SuSE, even Sun. Sun is a great server company, and so is Red Hat. But that business is very different from the desktop.

That's why we say until you can walk into a [large electronics chain store] and see a Linux computer on the shelf, desktop Linux will not go anywhere.

How will desktop Linux get into those chain stores?

One of the myths of open source is that, well, you just put the free software out there and then it ends up on computers. It just doesn't work that way. The original equipment manufacturers need someone they can call, and they need to have an economic incentive.

The source code might be freely floating around on the internet, but that doesn't mean it's free for the end consumer.

Everybody has been looking at supporting the user for Linux to get traction. You're saying the key is the manufacturer?

It's a combination of competitive pressures on retailers and us having a salesforce that will go out and meet with those folks and say: "We will support our product, we will certify our software to run well with your equipment."

End user support is Red Hat's model, and that works very well. But Red Hat's customer is very different from our customer. They are very technically savvy users. We are focusing on the value customers, and those budget-minded users are almost by definition not very technically savvy.

We focus on price. It's not about Linux or open source, it's not about a political statement. It's all about value.

In July you settled a legal dispute with Microsoft over the name Lindows, in which you received $20m (£11m). Before you named the company Lindows, it was called Lindoz, which carries less resemblance to the Microsoft trademark. Were you trying to get sued?

You're looking at [the Lindoz spelling] from a too technical standpoint. Why did Microsoft sue our company? Was it because they thought people were really confused between their products and our products? That's ridiculous. Every packaging we have says "We're not Microsoft".

I think we were sued because we threatened Microsoft's monopoly on the desktop. I know I'm competing with the richest company in the world. You have to expect that they are going to shoot every bullet they have at you. So I wasn't surprised, but I wouldn't say we planned for it.

In your office conference room there's a famous picture of the student standing in front of the column of tanks at Tiananmen Square. Does that symbolise Linspire's position towards Microsoft?

There is some analogy, but I remind people that this is real life. A lot of people got run over by tanks that day. This guy is risking his life. I'm not risking my life; I don't get killed if Linspire doesn't work. So it is overstated a little bit.

Do you think investors lost interest in the initial public offering?

The fact that we postponed our IPO doesn't reflect necessarily on our company. It's just a reflection on the current market. I don't think you can say "we did X this year, so that is what desktop Linux is going to be from this year on".

I know the sentiment is that unprofitable companies and early stage companies shouldn't go public. I disagree. To me, it's absurd to say that only companies that are profitable should go public. Then you are saying that only companies that don't need money should go public.