Mariano Blejman

Editor-in-chief Suplemento NO Cultura Digital Página12

  • Increase font size
  • Default font size
  • Decrease font size
  • XML Parsing Error at 1:32. Error 73: > required
Home Tecnología INTERVIEW EXCLUSIVE Linus Torvalds, creator of Linux, open operating system

INTERVIEW EXCLUSIVE Linus Torvalds, creator of Linux, open operating system

"Many patents are totally ridiculous"

The creator of Linux in two decades tells how his invention went from being the heart of an open operating system to become a friendly desktop environment and in the center of Android phones. "There shouldn't be a single ideology", he says. Published on May 30, 2011 in Página/12 (Original in spanish click here)

By Mariano Blejman
--They are marking 20 years of the birth of Linux, or what date do you think is what gives the birth to Linux?
--Well, for me it obviously wasn't any one particular date, since I was working on it for some time before it ever got released. At the same time, I do think any of the dates being bandied around are reasonable: the actual date when something was announced or released is obviously meaningful. So just depending on how you count, you can do three different dates. The one I personally think is "most relevant" is September 17 – that was when I actually made the linux-0.01 tar-ball and uploaded it to a public ftp-site ( However, I never actually publicly _announced_ the 0.01 release (I only sent emails to a few people in private), so for that reason two other dates tend to be mentioned too: October 5th was the first time I announced a Linux release publicly (the "Do you pine for the nice days of minix-1.1, when men were men and wrote their own device drivers?" announcement for Linux-0.02 on the minix newsgroup). And some people count July 3rd, because while I wasn't ready to publish anything back then, it's the date of my first public mention of working the project. So it's a matter of taste. Personally, I would tend to use the September 17th as the birthdate.
--Have you ever thought that Linux was going to be so big?
--Obviously not. At the same time, most of the growth was pretty gradual, so there was never much of a sense of huge surprise at any particular time. Only looking backwards do you get that feeling of "well, that worked much better than expected".
--Do you think that Linux has any political sense, any social contribution, or merit is simply productive?

--I think it has all of those issues, to different people. Personally, I did it (and still do it) for my own purely personal reasons: I think it's fun and interesting, and I wanted an operating system for my own use. The fact that other people have helped, and that they sometimes have different reasons for helping (ranging from purely making money to social reasons to political opinions) is interesting, but those reasons are still not why I do Linux. Of course, the fact that other people are enthusiastically involved, and the fact that Linux makes a difference for so many people, that does help motivate me too. I enjoy working on Linux for its own sake, but I obviously also enjoy the fact that it's a big project that has had an impact in the world.
--What do you feel having your name associated with a product used by millions of people around the world, even without knowing it?
--Heh. It's great, of course. We all want to feel relevant, and feel like we're making a difference in the world. Having a job where you feel productive, and know that the work you do *matters* is a big deal.
--Which is the state of Linux? Number of lines of code, number of people working...
--Number of people is hard to estimate. It's easy to give rough numbers ("about 1000 people are credited as authors in each kernel release in the source control logs"), but what does that _mean_? Some of those people do trivial one-liners, others write thousands of lines of code. And what about all the people who do testing and other support? As to lines of code: the current kernel source tree is about 14 million lines. Not all of that is *code*, obviously - that includes all the comments, the documentation, the build infrastructure, and some tools code too. Over half of it is drivers, a big chunk of it is architecture support for the 20+ architectures we support, and we have 60+ different filesystems of which most people only use one or two. So of the 14 million lines of kernel sources, a lot of it is features that do not affect most users. The really "core" kernel is much smaller. But you can really count the other way too: what is "Linux"? It's not necessarily about just the kernel, it's all the projects around it, some of which aren't even specific to Linux but are used on other operating systems too. So it really is hard to give one single number for anything.
--Which are the main challenges?
--For the kernel, one of the biggest issues is simply hardware support. Supporting all the random hardware out there is what most of the actual programming effort goes into. At the same time, we've had a lot of challenges at a maintenance level too - simply the issues of how to work together in a fairly loosely-knit community, building up the infrastructure (and just organizing the source code) to make it possible to work together. Some of it is tools (like the "git" project to actually maintain the source code), but much of it is a more ethereal "community" issue – just building the social links between people to make it possible to work together.
--Which are the main partners?
--Odd word choice. I have a lot of people I work with rather closely, and trust personally. They tend to work for various tech companies that are involved in Linux, but I work with them purely as people, not as "company representatives". So I trust them personally, not because they work for so-and-so company that works on some particular feature. Obviously, there are a lot of companies that end up being very instrumental in helping support Linux. They do different things, and tend to concentrate on different areas - and not all of it is about writing the code at all. In addition to emplying the engineers I work with, companies do marketing, they do bug triage, they do user
support.  It's all important. And I'm not going to name names of either the individuals or companies, because I'm not going to be able to say who is "more important" than others: it depends on your interests and your use.
--Wich is the main enemy?
-I really don't think that way.  I do Linux for my own positive reasons, and when I compare it against anything, it's against itself. I want to improve Linux to be better than it used to be, not compete against anybody else. So I used to make jokes about microsoft, but it really was never about them, or about any of the other  technology companies.
--When I talked about linux enemies: Aren't the "private patents" enemies of open source "movement"?
--Ahh, yes. Patents are a problem. Many patents are totally ridiculous, but fighting them is hard and costly. The good news is that most companies hate them too, so there's some hope that the system will change or at least improve a bit.
--What distribution of linux recommend and use?
--I personally happen to use Fedora, but the important word there is "happen" - it's largely from random historical reasons. I care about kernel programming, so to me a distribution is just a way to get a new machine to be useful, then I don't care that much any more because I will replace the parts I really care about deeply (which is mainly just the kernel, git, and historically some other projects if needed). And recommended distribution really ends up being about what your use case is. Is it android for phones, ubuntu for the low learning curve, or any of the other personalized distributions will depend on you. For most people out there, the best distribution to use ends up being the one that people around you use, so that you can share experiences and learn from each other.
--Do not you think that Ubuntu sometimes goes too fast, on upgrades and sometimes it is counterproductive?
--I don't think so. You want cutting-edge distributions trying out new things, the same way you want very stable distributions that stay at old versions for a long time because they don't want to rock the boat. As a technical person, I think the cutting-edge ones are much more interesting, of course. And for many users it's the right way to go too: you get early access to new features and capabilities. Of course, it does come with all the sharp edges that come from being shiny and new, so some people will definitely prefer a more staid approach.
--What desktop environment should be used?
--There is no "should". It's a matter of personal preference, and what you are used to. I've had some really bad experiences with desktop people who decide to change the world, so I switched away from KDE when they did their big KDE-4 change. And right now I'm staying far away from Gnome-3 for the same reason. The desktop, more than just about anything else, is about "used to it". That's obviously one reason why the desktop market in general is so hard to change.
--The term "open source" leaves the door open to get to the linux kernel proprietary software?
--No. "Open source" is very much about not being proprietary.  That's what the "open" is all about.
--What ideology has Linux?
-I don't think there is an ideology, and I don't think there *should* be an ideology. And the important part of that is the "an" - I think there can be *many* ideologies.  I do it for my own reasons, other people do it for _their_ own reasons. I think the world is a complicated place, and people are interesting and complicated animals that do things for complex reasons. And that's why I don't think there should be "an ideology". I think 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 - and they think that open source is a good way to do that. That's _one_ ideology. I think it's a great one. It isn't really why I started doing Linux myself, but it warms my heart to see Linux used that way. But I _also_ think that 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 that's a perfectly good ideology too. The world would be a _much_ worse place if we didn't have companies doing things for money. So the only ideology I really despise and dislike is the kind that is about exclusion of other ones. I despise people whose ideology is about "the one true ideology", and not following that particular set of moral guidelines is "evil" or "wrong". That's just small-minded and stupid, to me. So the important part about open source is not the ideology – it's just that everybody can use it for their own needs and for their own reasons. The copyright license is there to keep that openness alive, and to make sure that the project doesn't fragment into people who hide their improvements from each other and then have to re-implement each others changes - but it's not there to enforce some ideology.
--Does the international crisis has been an oportunity of growing for open source?
--I wouldn't really say that. I think that in some cases you have hard times showing you the reason to do something (the expression "necessity is the mother of invention" being about how need and hard times can be good ooportunities for new ideas and things). But at the same time, I actually think that most real development happens without a crisis. So now, in an economic downturn, many companies are turning to Linux and open source because you can avoid license fees etc. But at the
same time, I'm looking at the time before the crisis, and people were using Linux back then in new and exciting ways too.
--Do you think the Android phenomenon is another example of the power of free software?
--Absolutely. The notion that you can take open source software, and do things with it that weren't ever planned by the original people, and use it in new and surprising ways is very much part of the whole idea of open source. Android is a good example of how Linux - who most people thought of as a server operating system ten years ago - is now very much a cellphone OS too. And it's exactly because people could tinker away with it, and do their own thing.
--Do you think about the notebook Chromebook Google? It is an irony of open source software have made an open system while leaving the user as "slave" of a single company?
--You really have a very negative world-view, don't you?
--No, I don't have a negative world view, I am just journalist! :-)
--Hey.. Much of my family are journalists (mom, dad, uncle, grandfather..)
--But it is not ironic?
--I don't think you need to be pessimistic even as a journalist. I'm not sure where chrome will go, but at the same time it's very clear (just look at all the cellphones and tablets), that most non-techies really don't want a general-purpose "computer" – there seems to be a fairly large base of people who really don't want to maintain their own computer setup, but want to get access to the most common things - web browsing, email, some text processing, photo management etc. And while tablets may be sexy right now, I think a lot of people do want the keyboard and mouse. Writing stuff on a tablet really isn't all that great. So I think a chromebook makes sense in that kind of area. Why would that make anybody a "slave"? It's about convenience. Are you a slave to the electricity company just because you depend on them (and have to pay them) making electricity available to you?
--Does Openoffice and Libreoffice case demonstrates the strength of free software or open source and the "dictatorship" of communities, or is an exceptional case?
--I actually think OpenOffice is another example in  along string of patterns where people tried to "control" a project a bit too much, and it eventually broke down because the controlling party wasn't on the same page as the users. The move to Oracle and the tightening of that control was just the thing that broke it entirely, there had been grumblings for years about the way OpenOffice was developed. And no, I don't think it's exceptional at all. Many projects have gone through that kind of situation, and what ends up happening is thatwhen the problems become too acute, somebody forks the project. That's a big and painful step, and forks are not always successful, but they definitely happen. And sometimes the fork ends up being temporary, and it's an event that shows the original main developer group that they can't ignore the other pressures. In those cases the fork gets folded back in, and usually that also involves an opening up of the core group. And sometimes the fork becomes a wide rift that never closes, both because of technical reasons (the changes may have been too big to merge back) but mostly because the two projects have very different views on where they want to go. XEmacs vs GNU emacs is a fairly well-known historical example of that, but lots of projects have gone through that phase. And I think forks are a good thing. It's what keeps people honest in open source. Any maintainer of an open source project, knows that he/she need to keep an open mind, because otherwise somebody else can just come in and fork things. So forks can be very acrimonious and painful, but I think they are very much a part of the whole open source model.
--Linux intend to stay in the current GPL or plan to migrate to GPL 3?
--Oh, Linux will stay with GPLv2.
--What exactly is your daily work now? Keep writing code?
--I write very little code these days - I read email, and I merge other peoples code, and I discuss changes and tell people why I _won't_ merge their code. So 99% of what I do is about communication, and maintaining the central kernel source code repository - without really programming myself. Oh, I make changes too, and every release tends to have several commits written by me (in addition to the hundreds of merge commits I do), but it's not a big amount of code in any real sense.
--Why do you think has not been able to expand a good mobile platform independent as Meego?
--I think the whole mobility part is pretty new. Even just five years ago, most mobile embedded vendors didn't really want a big operating system: CPU's that were low-power enough to run in a truly mobile environment were also weak enough that most of them tended to run special embedded operating systems.  So Linux was strong in embedded before, but it tended to be in things like routers etc that weren't small battery-operated things. Things have obviously changed, and hardware (and usage cases) have changed, and these days you want a pretty full-fledged operating system even in a small hand-held device. But it takes a while for the infrastructure to sort itself out. Also, the embedded market itself has a lot of historical baggage, and the developers in that area have a long history of making one-off devices and not much of a high-level "platform". That's slowly changing, as they are starting to see that there is advantage to trying to commoditize more of the stack and the hardware vendors are more and more turning to standard generic SoC devices rather than the mixture of random chips they used to have. So the market is slowly changing and that will make it easier to have a common mobile platform, but we're still not quite there yet.
--When is going out the kernel version 3.x?
--I am actively considering making the next version 3.0, partly because of the whole 20-year anniversary, but also simply because the numbers are getting a bit big (the "2.6" part of the version number has been going on for so long that it has become almost meaningless, and the "39" part of the current version is just a random integer that isn't particularly easy to remember). (Note: one day after, Linus Torvalds accepted the kernel 3.0 rc)
--What are the hardware companies more reluctant to give support to Linux?
--Most hardware companies are actually getting pretty good at supporting Linux. But many of them just don't have good documentation (and, perhaps more importantly, don't have a _tradition_ of writing public documentation at all), and some of them are still kind of sitting on the fence. There are a few companies that seem actively reluctant. NVidia in the PC space has been a problem, as were historically wireless chip manufacturers. The wireless people seem to have gotten over it, but the graphic chip manufacturers seem to still be a problem, so now in the embedded space it's often problematic to get good 3D accelerated drivers.
--I supose that you are talking about IP as Intelectual Property not Internet Protocol?
--Yup. And I really don't know. It's one of the issues that has been quoted as a possible reason for both closed source code and closed hardware. Another reason (especially for closed source code) is just "it's ugly and buggy, and we're too embarrassed to let you see it". As to why? Who knows. Maybe they are afraid that they stole somebody's IP, and would get sued if they made it public.
--Last but not least, should you bothe sit down with Richard Stallman to iron out differences, or they are irreconcilable?
--Oh, I've met rms several times, and we have very different ideas of how things should be done. He is very much about the whole "one ideology" thing, and I'm very much against it.
--Why do you think increasingly not-using the "name" Gnu, when talking about Linux?
--I've never used the name GNU. Linux has never been a FSF project, and the FSF never had anything to do with it. Most of the tools aren't GNU tools either, although the GNU C compiler was (and is) a big deal. So the term "GNU/Linux" never made any sense. That said, I also never felt that people couldn't call it anything they wanted. Most of the distributions give the system their own name: Fedora, SuSE, Ubuntu, Android, Mandriva, the list goes on. So if the FSF wants to call it GNU/Linux, why should I care? It makes no more sense than calling it after a kind of head-wear, after all ;)

Twitter: @blejman






My Paper.Li!!!

Mariano Blejman

Editor-in-chief of Suplemento NO and Digital culture editor and writer at Página/12 newspaper, located in Buenos Aires, Argentina. Journalist. Web advisor. @HacksHackersBA organizer. Obsessively profiling and reviewing Internet.

View Mariano Blejman's profile on LinkedIn

Twitter @blejman
Twitter @supleno
Twitter @cult_digital
Twitter @HacksHackersBA

Hacks/Hackers Buenos Aires

Existe en Buenos Aires un lugar donde los periodistas y los programadores pueden juntarse: los periodistas (hacks) saben contar historias, pero muchas veces tienen difusos conocimientos tecnológicos. Los programadores (hackers) saben manejar datos digitales, pero a veces carecen de criterios informativos. La apertura de Buenos Aires de Hacks/Hackers (creado por Burt Herman de intenta acortar esa brecha: marcará agenda en materia de periodismo digital, funcionará como espacio de networking e influirá en los medios locales sobre las nuevas formas de hacer periodismo de cara al futuro.

Podés sumarte a HacksHackersBA, seguinos en Twitter, o lee lo que está pasando en el blog oficial.


Manu Chao frente a su condición de líder social

"No sé muy bien qué hacer con eso"


Atención amigo manuchaoista. El ansiado Próxima estación: Esperanza será editado finalmente en junio. Antes de eso, en la comodidad de su hábitat cotidiano (Barcelona, Barrio Gótico), el francés loco más querido en Latinoamérica se colgó un par de tardes, noches y trasnoches con el No. En la calle, en bares, en una cancha de fútbol, Manu habló de sus obsesiones, de sus canciones y de las revoluciones por venir. Original

Note: Since year 2000, I did about fifteen articles with Manu Chao between exclusive interviews and chronics, arround Argentine and Europe. Clicking here you'll see full list of links of articles related with Manu Chao. Some day, if he wants, a book will arrive...