Talking to Torvalds

Linus Torvalds He hates cellphones, but thinks that acceptance of the open source concept is now taken for granted - in a good way. BCS managing editor Brian Runciman interviewed Linus Torvalds after he received the BCS Lovelace Medal. This interview also appears in the ebook Leaders in Computing.

Let’s start by clearing this up: Lie-nux or Lynne-ux?
Well it is Lynne-ux, because I was working on Minix, it sounds similar.

Isn't your name pronounced Lie-nus - like Linus from Peanuts?
I was partly named after him, but in Finland it's pronounced Lee-nus anyway - so it doesn't work that way either!

BCS is pursuing professionalism in IT - what are your thoughts on this?
Well, I'm self-taught. The Linux community itself has been self-organising and most of the early people involved were students, not even programmers. Many were from a physics background, using Linux for calculations and physics problems. So I don't really have an opinion in this area.

Linux hasn't made huge inroads on the desktop - is that a worry?
Well the desktop is not really a huge market for Linux. It's hard to get hardware manufacturers to support it. They won't even tell us, for example, how their new graphics chips will work, so it's difficult to write code for them.

The desktop is special, there's a huge inertia for people because it's the only place we are really aware that we are using software. In things like iPods the software is almost invisible - it's updated behind the scenes, it only has a limited interface.

For Linux it's been easier to be in servers or embedded systems because all that matters is that it works - people don't care about what runs the system, only about the output. The desktop is a more direct interaction.

Looking back on 16 years or so of Linux is there anything you would do differently given the chance?
I think it's the 16th birthday in a couple of weeks. Realistically I wouldn't have done anything differently. A few technical choices weren't always right, but they are easily solved. But things like the community set-up and the licensing approach worked out really well, spot on.

Did Linux develop as you expected?
No, I expected it to be much smaller, just as something I would use - the commercial side wasn't planned at all. I did it for fun, because I was interested in software and interaction.

Now I do it mostly for the social side - I never thought of myself as sociable in that way, but it's fun to interact via email.

How people use Linux now was also unexpected. For example, it's used a lot in third world countries, although I'm not involved personally in that, and that's good.

BCS is celebrating its 50th anniversary this year. What developments in computing do you think are the most exciting in the last 50 years?
What I think about is IT's ubiquity. I don't like the word, but when I started out programming was viewed as odd and computing itself was expensive. But now interacting with computers is so much a part of people's everyday lives.

That has produced changes in the way we programme too. Because of the power of hardware now we do things that wouldn’t have been possible even five to 10 years ago - things that then would have required real high-end hardware to solve problems that would have seemed unrealistic. So the growth of computing power and its availability have been the big things for me.

There are concerns in the industry with getting people into IT in a serious way, not just superficial usage. How can we do that?
I have this problem with my daughters now. They use computers, but to make the leap to real interest in technology and programming, I don't know. My eldest daughter is doing a Lego robotics course this year, which involves a little programming. She uses Linux, but only to play games, surf the web and email.

Personally my interest started with a small computer club - there were only two of us, those who actually had computers. I had a Vic 20 and my friend a ZX81.

Who in the IT industry inspired you, or was a role model for you?
No-one from the industry. My role models were always scientists.

In fact I use this as an analogy for open source: it's science versus alchemy or witchcraft. I've always felt that the way to progress is to 'stand on the shoulders of giants' and that needs openness, not a form of witchcraft where how you do things is hidden, where you protect your knowledge.

My inspirations were people like Einstein and Newton.

What recent developments by others have impressed you most?
One of the people I'm a fan of is Richard Dawkins. He’s very famous in the US and UK for his anti-religious stance, but I've found his books on biology and genetics much more interesting. I find biochemistry fascinating.

Will computing and genetics ever truly overlap then? I'm thinking of Ray Kurzweil's views, for example?
I think the singularity idea is over-hyped. It's a cool vision, but whether it will happen in reality I'm not so sure. But both sides can give to each other.

To make AI really successful it requires an understanding of neural networks and how the brain works. And genetic research today is possible because of the processing power we now have. The protein folding stuff is very processing power intensive.

How do you feel about the media's approach to open source?
It's been very easy. Journalists tend to be very interested in communications and openness so I found it very easy to explain the philosophy behind open source. Most of my family are journalists so I've noticed how my interviews have changed over the years in this regard. I don't do many interviews; enough to see the changes, but not so many that I can't see how they’ve changed.

In the early days there was scepticism about why doing things openly would help or even work for programming and code. But now it's absolutely taken for granted that this is a good way of doing things. People know it's a great model. So the questions I get have changed because the assumptions have changed.

What are the biggest challenges open source faces in the near future?
I'm not that worried because it is just a better way to do things. People think of open source as a philosophy of freedom, but I think it's just a great way to get things done.

There are some worries, software patents are number one, but that doesn't just affect open source, that's a problem for all development. It's just a bit more obvious for open source.

What one piece of careers advice would you give to someone going into the industry?
What I saw a lot when I was at university was people getting into programming because they saw it as an up and coming area and a good opportunity to make money. It's the worst reason to get into anything, but especially programming, I think.

If you don't enjoy doing it you'll never be as good as someone who does. So you'll be second rate and unhappy.

That applies to anything. You need to like what you're doing.

Quick questions

Mac or PC?
Who cares? I have both, they both run Linux.

Are you a geek or a nerd?
I use geek, but I'll answer to nerd.

Smartphone, PDA or iPhone?
None, I hate cellphones, I find them really annoying - why are people disturbing me? I used to have one so that my wife could contact me, but I work from home now anyway.

How would you like to be remembered?
I think as someone who made a difference. A positive one, of course.

If not in IT, what would you be doing?
If I was in high school now I may well want to be a geneticist.

Linus Torvalds was awarded the BCS Lovelace Medal in 2000, and was presented it on his first visit to the UK in September 2007 by past-president David Hartley.

The citation:
Linus Torvalds created Linux, a remarkably efficient, stable and scalable UNIX-like implementation that is supplied free, over the internet in both binary and source formats. The significance of Torvalds' project is threefold. 

Firstly, the quality of the software engineering contributed by himself resulted in a robust and efficient operating system kernel now used by an estimated 7.5m people (as of 2000).

Although the majority of Linux users host the operating system using PC hardware it has been implemented on machines as diverse as palm pilots and massively parallel networks of microprocessors.

Secondly, Torvalds demonstrated that the very best software, including operating systems, is not (and perhaps even cannot be) developed by large corporations intent on profit.

Thirdly he inspired a large and growing number of developers to contribute and cooperate on what is one of the most ambitious distributed software development projects to date.

September 2007

Read more BCS interviews