Behind the Linus curtain


I've wanted to interview Linus Torvalds for some time, but with the holiday season, his new baby, and the release of 2.4 already on his plate, I knew my chances were slim at best. Timing is everything, it seems. Now that things are somewhat back to "normal," whatever that is, I tried a direct approach and appealed to his finer instincts. I've owed him a beer for almost 2 years now, so I offered him a case of Guinness in exchange for an interview. It worked. After predicting that you would pass the kernel development baton to Alan Cox this year, I read an interview of yours in which you put that notion to rest, attributing such speculation to people mistaking your humor for burnout. I admit that's where I went wrong. But along the same line, is there any sort of plan for how Linux kernel development would proceed if you were kidnapped by aliens from outer space, or otherwise became unavailable?

Linus Torvalds: Oh, there's a lot of confusion here, probably because a lot of people get so hung up about "ownership transfer," when I personally don't think such a notion even exists in Linux.

I will hand over the baton to Alan Cox this year, the same way I did for 2.0.x and 2.2.x -- he's really good at maintaining the stable kernels, he seems to enjoy it, and people trust him.

My handing 2.4.x over to Alan Cox doesn't mean that I step down -- it's just me knowing that what I like most is the development kernels, and while I revel in 2.5.x with new features, etc., I don't have the time or the inclination to also maintain the stable kernel at the same time.

This, I think, is what people have a hard time realizing. It's not about one person, and I've never done everything (well, strictly speaking I did do everything in Linux during most of 1991, but I got out of that "Linus is in charge of everything" mentality at the first moment I humanly could).

So I'm likely to never "pass the baton" in the way people expect -- I'm not the CEO (chief executive officer) who gives up leadership. I'm more the technical lead who concentrates on one thing, and it has just happened to be the most visible thing. The way the leadership has evolved, and will continue to evolve, is that others handle other issues -- to the point where I'm already just one of many. I'm just the most visible one, and the one whose decisions most people tend to respect. What do you most want to see in the next major release of Linux, and why?

Linus Torvalds: There's a few things queued up, the most noticeable of them being to finish up the ACPI stuff (which is in 2.4.x, but in an experimental stage), doing major real clustering and NUMA support, and just seriously revisiting some of the I/O path issues. Most of them tend to be fairly well-defined goals, which makes things easier.

Of course, the real fun is in what surprising things come up. Which, by definition, I won't know yet. As I've read over the archived email debate between you and Tanenbaum on the relative merits of the Minix and Linux designs, I'm struck by how easily you can express strong opinions without making it personal. Certainly, leading a diverse group of brilliant hackers from around the globe to a successful release time and time again is a Herculean task. I know that you claim to be "a bastard" and not to care what other people think or want, but don't you think your ability to lead developers is at least as important in the history of Linux as your code?

Linus Torvalds: I think one thing I do pretty well is not taking myself too seriously.

I seldom get self-righteous, and even when I am being impolite (almost always on purpose -- there's an art to insulting people, too), I tend to try to not be too serious about it. And most of the time it means that I can take criticism constructively, and sometimes just change my opinion on the fly and laugh at myself over having turned on a dime.

Of course, that's not always true. I've alienated my share of people too. I've gotten too upset for my own good on some issues. But on the whole, I can mostly laugh at myself and this whole mess called "Linux developers," which means that I get along with most people and most people get along with me.

And that, after all, is the point of being a leader. Ten years from now, what do you want to be doing (or not doing) and where do you think Linux will be then?

Linus Torvalds: You're asking me way too hard questions. I honestly have no idea.

I didn't have any idea 10 years ago when I started, so I'm convinced that not having a clue about where we'll be in another 10 years is a good thing. I'm not limiting myself with any artificial goals. Raising a family (congratulations on the latest!), working at Transmeta, leading the Linux revolution, and attending to dreary chores like this interview, for which I have pledged an entire case of Guinness, how in the world do you have time to do it all? Do you ever read, watch movies, or go to the park with your kids? Do you consider the Silicon Valley home now?

Linus Torvalds: I do consider SV to be home, and sure, I have time for other things than just work. We get a babysitter and go out for a movie every week, and I try to take a few days off every once in a while just to lie down with some not-so-literary work of horror/science fiction/fantasy or whatever. And my pool playing is steadily improving.

The secret to not getting burned out is to play at working hard, and not taking things too seriously. See above. Have you had any heroes or idols -- in software or any other fields -- along the way? If yes, who were they and what did you admire about them?

Linus Torvalds: Few heroes, I'm afraid. In computers, Kirk McKusick and Steve Wozniak are kind of heroes. But on the whole, I think I've cared less about "how can I be more like Xxxx" than about just trying to do what I think is fun, interesting, and right. Motto in life: "Do unto others what you would like others to do unto you. And have fun doing it." Looking back, has Linux turned out the way you hoped 10 years ago? If not, what would you change to make it do so?

Linus Torvalds: It turned out much more interesting, and much more fun than I could ever have imagined. I can't imagine anything that I could have done better, even in hindsight. Luckiest bastard alive, that's me.

You owe me a case of beer. Sucka'.

Free Course: JavaScript: The Good Parts
View Comments
You Might Like
Join the discussion
Be the first to comment on this article. Our Commenting Policies