Linus Torvalds interview: Linux and the future of computing

By Graeme Burton
17 Dec 2012 View Comments

For example, the merge window for our next release (which will be 3.4) opened just two days ago, so I have tons of requests to merge new code from the various subsystem maintainers. So I read the explanations of what's going on, sometimes ask for clarifications, and then actually merge the code into the "mainline" kernel, which is what I still maintain.

Further reading

But just so that it's very obvious: I'm definitely not alone, and that's obviously part of the whole point about open source. There are literally thousands of people involved in writing code for the kernel. Each of our releases (and we do about four releases a year) tends to have about a thousand people involved with actual coding - many of them admittedly only making tiny trivial changes, but that's how people get started.

So I end up doing the release management and working with people and integrating things, but I don't do a whole lot of coding any more. For example, of the over half a million lines of new and changed code that I have merged so far during this merge window, I think I personally wrote a little over a hundred. So I really don't consider myself much of a coder any more, I'm definitely "management", even if that's a dirty word to most engineers...

Q. How do you see your future?

I really enjoy my work and I don't really see that changing a lot. Of course, who knows what will happen in five to ten years? Quite frankly, though, if I'm doing largely the same thing, I won't be unhappy. Of course, the technical problems constantly evolve, so it's not really the same old thing, but that's what keeps it interesting.

I'm simply not a very flighty person. I think that's one of the reasons behind the success of Linux: I'm still doing it more than 20 years later and, while I didn't originally believe it would be that kind of "lifetime project", at the same time I don't really see myself suddenly deciding to do something else.

That said, outside of my work, there's obviously other things going on too. My kids are growing up, with the eldest in high school right now, and we'll be starting to worry about colleges etc. And I have my hobbies (mainly scuba diving), and if there is one particular "personal ambition" I have it really is to just have a "good life". Part of that is interesting and meaningful work, part of it is family, but part of it is just enjoying things like my diving too.

Q. What other third parties are driving these ambitions forward?

Absolutely. Much of the drive behind new code and features in Linux comes from commercial companies. That can be anything from hardware companies wanting to make sure that Linux works on the new hardware they are building, to support and service companies wanting to make sure Linux works as well as possible for their customers. And of course technology companies that have some new niche that they are going for, and that need some particular feature to really well.

In fact, my personal wants and needs stopped driving Linux development long ago. Linux has done everything I really needed it to do for a long time... Going back ten or 15 years ago, it was mostly individual, technically minded developers and some university research projects.

These days, it really tends to be fairly big technology companies that have a particular area that they are interested in improving and driving Linux forward in. What's so interesting - and so good - is how open source really means that all these different interests can "meld", and the end result is really much better for it. Exactly because there is not just one company with one single vision, but lots of people with different ideas, and that balances the end result out.

Q. Bearing in mind the vast global changes we saw brought about by technology in the 20th century, could similar leaps be made in the 21st century? Or do you see a comparative slowing down in the rate of innovation?

I think we're actually approaching a fairly interesting decade or two. We're not far from starting to bump up against some fundamental physical scale-barriers in making computing elements smaller and cheaper. It hasn't stopped yet, but electronics manufacturing is starting to get to the point where the feature sizes aren't that many atoms across.

Reader comments
blog comments powered by Disqus
Windows 10 - will you upgrade?

Microsoft has made an early version of Windows 10 - its next operating system - available for download. The OS promises better integration and harmonisation across platforms, including mobile and desktop. Will your business be upgrading?

36 %
31 %
13 %
20 %