An IIPM Initiative
Tuesday, January 26, 2021

"Linux doesn't have an ideology... and I don't think it should"


B&E | New Delhi, February 22, 2012 18:10
Tags : Linux | Linux operating system | Linus Benedict Torvalds | interview of Linus Benedict Torvalds | |


With a 61% market share of the global servers market, and running 75% of stock exchanges worldwide, the Linux operating system today powers the servers that delivers amazon, facebook, twitter, ebay and google. Linus Benedict Torvalds, the helsinki-born creator of the OS, speaks out his mind on the merits of open source, open collaboration software systems.
Q: In 1991 you were a student at the University of Helsinki and a self-taught hacker. What got you thinking about creating a new Operating System?
A: It wasn’t really a conscious decision; it was more a confluence of factors. Part of it was simply that I was interested in Operating Systems and had been working on low-level issues for a long time. I’d been doing assembly language programming and messing around with device drivers with my previous machine – a Sinclair QL that had very little support in Finland. So although I was only 21, I had something of a background for it. Another thing was that I wanted to run Unix on my newly-acquired PC, so rather than running DOS and Windows, I had gotten Minix for my machine, which was a small Unix-like OS built for educational purposes. But it was much more limited than the Unix I had gotten used to at university. At the same time, I was working on a ‘pet project’ to teach myself all about the innards of my new machine. This is what ended up expanding to become the first version of Linux.
Q: Linux doesn’t seem to have an ideology – or does it?
A: Linux doesn’t have an ideology, no,and I don’t think it should. The important part of the question is the word ‘an’; I do think there can be many ideologies: I do it for my own reasons, other people do it for their own reasons. It’s really refreshing to see people working on Linux because they believe they can make the world a better place by spreading technology and making it available to people more widely. That’s one ideology, and I think it’s a great one. It isn’t really why I started Linux myself, but it warms my heart to see it used that way. But I also think it’s great to see all the commercial companies that use open source simply because it’s good for business. That’s a totally different ideology, and I think it’s a perfectly good one, too. The only ideology I really despise is the kind that is about exclusion of other ideologies. That’s just small minded and stupid. So the important part about open source is not the ideology – it’s just that everybody can use it for their needs and reasons.
Q: Before long, you began to encourage input to your system’s coding from other members of the IT community. Given how hard you worked on it, how did feel about the ‘loss of control’?
A: To me, inviting other people to become part of the project wasn’t about me losing control; it was about getting lots of new ideas for further improvements. I would almost certainly have become bored with Linux rather quickly if it hadn’t been for this decision – that’s what had happened with the earlier projects I worked on in private. In fact, the initial impetus for making the Linux source code available publicly was not because I wanted others to help me write it – it was because I was proud of what I had done and wanted feedback on where to go next. The early interactions were less about other people writing code, and more about asking others what they thought the project needed, and then me writing the code myself. When people started actually sending me suggested code changes, that became a very natural extension of it.
Q: Some people have said that Linux is even more interesting from a social standpoint than from a technical standpoint. Do you agree?
A: I don’t disagree, but I think the technical side has been very interesting, too. Not because Linux is a radical new product (which it isn’t), but because the technology is exciting in and of itself, and that’s why we have attracted so many developers. Having said that, what was really new about Linux was the social development model. Linux was not by any means the first open-source project, but it was the first large-scale one where development was so widely spread out and open. Most projects at the time were fairly tightly-controlled and consisted of a group of people who met together physically. In contrast, from the beginning, Linux was all about email interactions between people who didn’t know each other otherwise. As mentioned, I never really felt like I had to control the end result. Sure, I used my discretion and would not apply just any random patch of code that came my way, but at the same time, from very early on the project was fundamentally about accepting not just new code, but new directions and ideas from the outside.
Q: Your company was one of the first to embrace an open environment. What lessons about its merits can you share with us?
A: One key lesson is to not try to control the end result too much. A fair number of open-source projects have been ‘technically’ open source, but the project leadership really acted as if the whole point was to generate a return for the originating group. If you do that, you are missing the whole point, and you are also going to miss out on the talents of the wider community. You won’t get access to people who are deeply committed to it.
Q: You believe that centralised systems never work as well as ‘distributed’ environments. Can you elaborate?
A: The kind of centralised planning that you so often see is a fundamentally flawed approach. It needs to evolve with very close feedback from users, and that cannot possibly reach all the way back to some central design person or group. I also think that any centralised system will be biased towards a goal. That can be beneficial if the goal is well-formulated and understood, because you can be quite efficient. But most real-world problems aren’t simple enough to be that well understood, even for a single use-case. Furthermore, few of today’s problems are of that ‘single use’ type; you always end up having different users that want widely separate things. And in that case, if you have a central core group that sets the direction for the project, it will inevitably end up being biased towards a particular problem space, and thus biased against some others. In the end, centralised design actually doesn’t work outside of trivially-simple cases. I also think that centralisation is bad in a purely technical sense. 
Q: Lastly, Linux on mobile devices has come a long way in the past two years, mainly due to the Android OS. Does it please you to know that Linux is in the hands of millions of people every day?
A: Android is a great example of how Linux – which most people thought of as a server OS ten years ago – is now very much a cellphone OS, too. And this happened exactly because people were able to tinker away with it and do their own thing. The thing that is the most fun for me is when people use Linux in ways that I never intended it to be used! 


Rate this article:
Bad Good    
Current Rating 5.0
Post CommentsPost Comments

Issue Dated: Feb 5, 2017