Submitted by “Linus Torvalds was only 22 when he decided in 1991 to share with friends and colleagues the code of Linux, the new OS he had created. The computer science student at the University of Helsinki could not imagine the revolution his decision would cause through the IT industry in the years to come. In this interview, he talks about why he released the code, offers his views on Microsoft and says the future belongs to open source.”
Same old story, same procedure as every year :o)
Well it’s not like he’s going to confess Microsoft is “the future” (whatever that means) after pouring his life for the past decade or so into Linux kernel and open source development.
We all had to expect such a point of view from someone so heavily invested in Linux.
No surprises here.
Its not really that important what becomes of Linux. If it were to eventually falter because of complexity there will be other Open Source OSes to take it’s place.
Whats is assured is that Open Source will continue to thrive long into the future, until “software” becomes obsolete. Perhaps Open Source will be used in alternate context thereafter.
Or there would be two people as rich as Bill Gates in the world.
Forget it. Linux wouldn’t be used by anybody.
* It’s main feature was that it was free. That was the only reason why it got attention and support.
* Without GNU, Linux wouldn’t be usable. The kernel alone wouldn’t sell to anybody.
As compared to today…?
Linus back then had skill and drive. If he had had initiative, too, we’d be talking about one of the richest men in the world right now.
Linux made the GNU project relevant, not the other way around.
And you severely underestimate what Linux’s commercial value would have been. Software we take for granted as free nowadays was selling for $50-60 back when he started.
As compared to today…?
What do you want to say? Linux is used by lots of people now. It wouldn’t be used then. There was enough competition.
Why should a half-assed kernel written by a single developer achieve what IBM didn’t with a complete state-of-the-art operating system?
Linus back then had skill and drive. If he had had initiative, too, we’d be talking about one of the richest men in the world right now.
Thanks for repeating yourself again. This is just a claim without essence.
Linux made the GNU project relevant, not the other way around.
This assumption is deadly wrong. The first milestone for Linux was that several GNU tools could be run onto it. Without them, Linux would just be useless for any productive environment. What do you do with an OS kernel without OS? Nothing.
And you severely underestimate what Linux’s commercial value would have been. Software we take for granted as free nowadays was selling for $50-60 back when he started.
What is the value of a hobby kernel, small part of the whole OS needed? There were lots others out there, and even complete operating system, with more power and a heck of more value.
The key point in Linux was, as Linus states himself, the attraction to other developers as a free operating system kernel, fitting into an already existing free operating system w/o kernel.
And btw, the most of Linux’ code is _not_ written by Linus. And you believe he could have developed a whole operating system and make the big money. That’s what several others tried, and eventually failed. There were lots of commercial unices out there, they all there beaten _only_ by price & free as in freedom.
Go ask Bill.
Bill never wrote a kernel. Perhaps do some history lessons and then come back.
No, but he wrote a BASIC interpreter that were used on many small machines, not that different from a kernel really – both did know how to program.
I know how to program too, but I wouldn’t compare myself to Linus or BG. A BASIC interpreter is quite a bit different than a kernel – there are some superficial similarities, but they’re really quite different.
>A BASIC interpreter is quite a bit different than a kernel
What kernel in the beginning? You cannot even compare the first “kernel” with the later kernels.
Developing a basic on this machine,
http://en.wikipedia.org/wiki/Altair_8800
was even more difficult than coding some “first-step-wannabe-kernel” I think.
“Programming the Altair was an extremely tedious process where one toggled the switches to positions corresponding to an 8080 opcode, then used a special switch to enter the code into the machine’s memory, and then repeated this step until all the opcodes of a presumably complete and correct program was in place.”
Try this and forget about nice editors and compilers.
You suggested that back then, a single person with a simple OS couldn’t compete with a large corporation with a full OS. I alluded to the case of MS-DOS, in which a technically inferior in its first versions but still functional OS, with smart planning and good business moves, which was still able to outmaneuver a giant.
Linux never was an OS, even not a simple one. And also, Linux was late compared to MS-DOS. Don’t you see that much more was needed to get into the market? Others had much more, but failed anyways.
Let’s be serious. The GNU project was as much a grassroots project as linux was when it was launched. GNU was working on hurd at the time, which utterly failed, even though Linus himself figured linux was going to be a stop gap to Hurd.
It was the commercial potential of linux that drove the commercial investment in GNU, which is just as dependant upon commercial contributions as the kernel itself is.
So, sure, it’s easy to argue that linux couldn’t have existed with GNU, but it’s also likely that GNU would barely exist without linux. Linux did make GNU relevant. Linus could always have figured out a way to work with a BSD toolchain, but he didn’t. GNU, on the other hand, never managed to produce a kernel, but they are now inextricably linked to linux.
Let’s just call it a draw and let it be.
No. The point is, he’s just an ordinary motherfücker like any other in the army of free software writers. Yeah, he carved his name on a big tree, and have other spots on his karma, but he should not be praised for that.
On the other hand, is there a physicist that really admired Einstein? Another fairy tale in the making…
Let’s just call it a draw and let it be.
i agree, anyway is not going to matter much when SUN cuddles up to RMS and FSF with the Opensolaris kernel, when SUN releases it under the GPLv3.
GUN/Opensolaris OS
Edited 2007-08-11 22:09
A few remarks:
1. There were other UNIX-like system around, and some were better and more complete than Linux. If Linux wasn’t free, customers would not pay attention. If it was very low priced, but proprietary, well, that might have worked, maybe. The price could have gradually increased, as the system grew more attractive.
2. There is nothing illegal about selling proprietary kernel with GNU tools. Kernel is independent software, runs in privileged mode, while the rest of the OS runs in a sort of lightweight virtual machine. I think that it is glibc that makes GNU tools work with Linux. Glibc can be replaced with other libc, and actually it was libc once.
3. There is a BSD version of programs known as “GNU tools”. These are, in fact, traditional UNIX command line utilities that existed long before GNU.
‘1. There were other UNIX-like system around, and some were better and more complete than Linux. If Linux wasn’t free, customers would not pay attention. If it was very low priced, but proprietary, well, that might have worked, maybe. The price could have gradually increased, as the system grew more attractive.’
Actually I do think that this is as far besides the point as you can get:
Version 0.01 sure as hell didn’t have any customers no matter how much it costs because it was hardly useable.
The difference that open source made was that developers got interested and worked with it.
If it would have been closed source it would have never went anywhere since – as was correctly pointed out – even the messiah of Linux (not open source – that would be RMS) couldn’t write a kernel like they have today by himself.
That the nixes got killed by it is more like a testemony to the incapability to the involved companies and has nothing to do with money.
After all it seems Apple is wildly profitable.
“Linus back then had skill and drive. If he had had initiative, too, we’d be talking about one of the richest men in the world right now.”
Why? I’m not saying your wrong about the “would have should have”, but we see from large companies the people that make the money are rarely the people getting their hands dirty with coding, although I can assure you that Linus is rather well off.
“And you severely underestimate what Linux’s commercial value would have been. Software we take for granted as free nowadays was selling for $50-60 back when he started.”
And compilers sold for thousands of pounds…whats you point.
I’d argue its a symbiotic relationship.
My point was that, back when he started, ‘just a kernal’ was quite a lot indeed; Linus was sitting on a gold mine. Heck, he was sitting on a gold mine factory. Things could have turned out much better for him and for Linux as a whole had he made it closed-source.
Hope that clears it up!
But you’re still missing many of the points he made in his article. He had no desire to run a business and didn’t believe he could do it successfully if had tried. As a proprietary OS, Linux would have been an also-ran that, even if it had been technically solid, would have been a blip in the marketplace. Microsoft’s economies of scale grew too strong for other OSs to really thrive. By divorcing Linux from any single commerical interest, including his own, Linus found a way for his new operating system to thrive despite the odds.
And then there’s the whole issue of whether Linux would have been as technically interesting as a commerical product. Linus, though talented, couldn’t have done it on his own, and it isn’t likely he could have attracted the world class talent that helped make Linux a success if he’d been looking to recruit them to work in Finland.
(No offence to Finland; it’s just hard to centralize talent.)
Edited 2007-08-10 23:17
There are many other kernel beside Linux.
Linux become popular because
1) it use GNU copyleft license and it is free
2) it use GNU tools
3) GNU HURD is nowhere in sight and everybody are tired of waiting.
4) the author of Minix don’t want too many features that would make it compicated
If Linux cost $$$ per copy. Nobody would want to use it.
1) Yes, being commerical would preclude being free. That’s kind of the point.
2) At that time, Linus was writing his own version of the GNU tools — and necessity drove him to write other key points of Linux’s functionality. It would still be many years before Linux was anything but a one-man show — and at that point, he could have started hiring people just as easily as he began ‘borrowing’ other people’s code under the GPL.
3 and 4) The inability of any other competitors (copyleft or copyright) to match what he was doing are both reasons that a commercial Linux would have had more than a chance.
And as to nobody wanting to use it? If it does what people want and need, people will use it. Linus GPLing his kernal had virtually no impact in whether or not people started using it.
>Software we take for granted as free nowadays was selling for $50-60 back when he started.
Nonsense, BSD was available too. And free software under the GPL was e.g. distributed back on the Amiga. It was maybe different in *your* very own context, but not for everyone!
And a lot of the software we are paying for to do was something we got for free back then
He DID have initiative – he just didn’t use it to get filthy rich. But he very much took initiative to push FLOSS to new speeds.
If he had had initiative, too, we’d be talking about one of the richest men in the world right now.
I think you misspelled “greed”.
Actually, I’d say they were mutually dependent. Linux needed GNU as much as GNU needed Linux. A userland without a kernel and a kernel without a userland are both utterly useless, but put them together and you have an OS. I don’t think either project would have gained nearly as much momentum alone as they did together.
BSD was an option too as userland, but like so many developers he made false assumptions about the lawsuit or didn’t know anything about it in the beginning of his development.
I’d say it’s rather been a symbiosis.
Yeh, GNU needed Linux as much as Linux needed GNU.
yep, for at that point they had the tools but no kernel to run them on. BSD was in litigation hell iirc, and hurd was nowhere to be seen (see linus’s comment on that in his first linux announcement. something about it could be here tomorrow, or it could be here in 10 years iirc).
Linux made the GNU project relevant, not the other way around.
This is one of the most fallacious statements I’ve seen in a while.
Linux was nothing without GNU – GNU/Linux was the basis for a working operating system that evolved into the global phenomenon it is today. GNU tools are still a core part of any Linux distribution, although the brand GNU has been marginalised by the brand Linux.
If Linux was not around, many people who contributed to Linux would probably have contributed to the HURD or one of the BSDs or one of the many other open source OS efforts around. The talent pool of willing contributors would more-or-less still be there, just distributed amongst different projects.
Of course the software landscape would have been different, but it wouldn’t have been that different. Linux is not some amazing piece of kernel genius. It’s good, but it’s replaceable given enough development effort, only it has absorbed a lot of developer hours that otherwise could have propelled one of it’s alternatives.
I dare say Linux is not the best technical solution out there as an OS kernel. It’s simply the most pragmatic, although lately such talent has worked on it that it has become a feat of engineering in some areas, but that could have or has happened to many the OS kernel alternatives.
Forget it. Linux wouldn’t be used by anybody.
Putting forth unfalsifiable statements is a preacher’s duty.
* Without GNU, Linux wouldn’t be usable. The kernel alone wouldn’t sell to anybody.
if gnu had not been available, something else would have. but as gnu was available, noone else bothered to provide a similar toolkit.
hmm, i didnt know the score of a post could go above 5…
Or 2 people half as rich or… in any case, the computer landscape probably wouldn’t be as interesting as it is now (and getting more interesting by the day
)
Very true
That is a typical excuse from someone who now, he could have made a LOT of money, but made the wrong choice, and is poor compared to what he could have been.
And off course he has a lot of excuses trying to make it sound like he is perfectly happy, making some remarks about roses and a poor deer.
Edited 2007-08-11 15:36
While currently monetary wealth is probably more often worshipped as any single deity, it is, for quite a lot of people, not the single most important goal to reach.
That is very true.
But here we don’t talk about having a little more money, but to have a huge amount, and NEVER EVER having to worry about money.
There is a HUGE difference.
Remember this is not just Linus not getting the job that would have gotten him twice the salary, it is make a choice that made sure he did not earn billion of $’s
There are things no amount of money can buy.
Absolutely. The difference is getting pies thrown at you, needing a body guard around whenever leaving the house, needing professional protection for your family, never ever again be able to do what you like because it might effect your company’s image.
I’d rather be able to negotiate any salary I want, be able to work from any place in the world (not needing to show up in an office every day), be free to replace anything if a better implementation arrives, …
The fact that Torvalds thought that the biases revealed by the arguments over the Novell/MS deal were more interesting than the deal itself was a rather interesting bit. He usually seems to be for things (like good technology) instead of being against things (like all things Microsoft). A such, I generally find his comments more refreshing than the vast number of people that are always pointing out the negative side of things. He’s had plenty of negative comments before, but overall, he seems to look at the positive side of things.
He’s always been a pragmatist who only gets hot in the head when the technology, not the ideals, are compromised.
Then again, I kinda find Stallman to be a pragmatist in his own right, just more passionate about the ideals than the technology. He’s not opposed as much to Microsoft, or against single companies, but more or less against the existence of closed-source software, and the very idea of it, in its entirety (the equal-opportunity activist, I suppose).
So I think they complement each other: the greasy engineer and the bearded academist. The fanboys and “advocates”, OTOH, don’t do any good to that relationship: they’re emotionally driven to hate a single group of people (“die Bill, die!”, stab stab…), rather than appreciate and make use of what makes FOSS the attractive model of development that it is.
I don’t get such vibes from either Torvalds or Stallman. They say what they need to say (good or bad) and then go back into pursuing their respective crafts.
“He’s always been a pragmatist.”
Could you give examples of this other than his use of Binary blobs in the kernel.
How about the fact that he has claimed, ad nauseum, that he chose GPL v2 for “pragmatic” reasons. Frankly, Linus is criticized for being pragmatic so often, I’ve never had to see someone actually ask for citations.
Google is your friend.
He cares about technology, nothing more, nothing less. He isn’t found of the faith of some GPL zealots. So he is a pragmatist, no believer.
He cares about technology, nothing more, nothing less. He isn’t found of the faith of some GPL zealots. So he is a pragmatist, no believer.
Nope. He’s a technology zealot.
Yes, and he deprived us of plenty good flaming Shame on him !
It’s very obvious CW tried to trap Linus in a flame of war between Linux and other things. (MS, GPL3…etc).
Then they can create some shocking news:
“Linus said, ………..”.
And it is also obvious that Linus is so careful of his words, and he is tune was totally different from when he is in LKML. Linus is sensitive to these traps.
…of the highest caliber.
Contrafactory history is a joke. Free your minds of what “could’ve been” and leave room for “what will come”.
Damn… if I’d only rolled a six that time..
Edited 2007-08-11 00:04
there is some really good stuff in there.
first off finding some project to work on when it comes to programming.
second, the talk about inertia. even ms gets hit by its own inertia now with vista.
and if your going to, as a corp or similar, fight inertia in your upgrade cycle. you can just as well future-proof by jumping onto something that seems unable to grow stale.
hell, if anything should be a eyeopener for corps it should be dell showing off a laptop running windows virtualized with no other cost then the hardware and the windows licence.
the only thing MS could do against that would be to put a limit on what virtualization software windows can be used with. but that would show their hand, and be anti-trust bait like nothing else
ng else 
Browser: Mozilla/4.0 (compatible; MSIE 6.0; ; Linux armv5tejl; U) Opera 8.02 [en_US] Maemo browser 0.4.34 N770/SU-18
there is some really good stuff in there.
first off finding some project to work on when it comes to programming.
second, the talk about inertia. even ms gets hit by its own inertia now with vista.
and if your going to, as a corp or similar, fight inertia in your upgrade cycle. you can just as well future-proof by jumping onto something that seems unable to grow stale.
hell, if anything should be a eyeopener for corps it should be dell showing off a laptop running windows virtualized with no other cost then the hardware and the windows licence.
the only thing MS could do against that would be to put a limit on what virtualization software windows can be used with. but that would show their hand, and be anti-trust bait like nothing else
ng else 
Browser: Mozilla/4.0 (compatible; MSIE 6.0; ; Linux armv5tejl; U) Opera 8.02 [en_US] Maemo browser 0.4.34 N770/SU-18
Is Opera now considered a second-class citizen on OSAlert just because it’s running on a handheld?
The Nokia 770 has an 800×480 screen, guys. No need to cripple the site just for it…
Browser: Mozilla/4.0 (compatible; MSIE 6.0; ; Linux armv5tejl; U) Opera 8.02 [en_US] Maemo browser 0.4.34 N770/SU-18
…are you yapping about?
“I don’t see the kernel changing licenses”
There it is folks! Batten down the hatches, our path is set!
I enjoyed the interview. Linus was rather muted; I prefer it when he is an arrogant, opinionated bastard.
Linux will continue to develop and grow. As Linus said the market place will dictate what happens with software, not FUD. In my opinion I think Linux is starting to gain traction in 07.
Interesting times ahead for my favourite OS:-)
I think the reason linux is where it is now is because of timing. Linux came out just as the internet was growing into a tool to allow collaborative programming. A new community based project comes out just as a medium to allow open information dissemination arrives on the scene. I think its just pure luck things happened when they did. I think it could have just as easily been BeOS if Be had made the OS a community project. But whatever, I am happy with open source and I will never go back to a closed environment.
I bet linus probably thinks the biggest threat to open source is the GPL v 3 license . there is probably no surer way of killing linux then somehow making a grave error with it. If they cut out commercial companies for instance it would be bad.
“I bet linus probably thinks the biggest threat to open source is the GPL v 3 license ”
I bet you haven’t read the interview. The bottom line is linus is only concerned with the kernel, and is very vocal about what he wants with his *code*.
This really has been done to death, but the bottom line is its the same license with much of the abuses cut out. The reality is developers have to choose GPL3 or GPL2 or BSD or whatever, and they think that is best for there product, personally I cannot see *any* benefit from a company abusing GPL, so why would I care if they are cut out.
From the interview, which you apparently didn’t read:
Again, it’s not that big of a deal — we have something like 50 different open-source licenses, and in the end, the GPLv3 is just another one. I don’t use the BSD license either, but tons of other projects do. Whatever suits you.
I think history has shown that Linus isn’t shy about voicing his opinions, even if they’re controversial. It certainly doesn’t sound to me like he thinks the GPLv3 license is the biggest threat to open source software.
He never has…thats just anti-gpl people taking him out of context.
Linus basically just says GPLv3 is all fine but he doesnt see it neccesary for Linux kernel to adapt to it while GPLv2 works great for his and many other kernel devs purposes
Stallman worked for many years (decades?) to provide a free Unix clone. Everything was done, except for the kernel. Then comes Linus and writes a basic, nothing fancy, kernel and get the credit? What would you think if you built a house from the ground up, and suddenly someone comes in and puts furnitures there and gets the credit? The house is his, and built by him? I wouldnt like it. And Linus doesnt want to give Stallman enough credit too. Just read what he says.
And Hurd, it takes long time because theyre basically doing research. You can never time estimate research. Theyre implementing state of the art things. If they wanted they could have done a basic kernel a la Linux I am sure that a few kernel developers together could have done that, in shorter time than a lone 22 year old guy. Linus also says that his code then, was pretty bad. He was a bad coder, then? Hurd is very diligent and has high ambitions. Stallman wants it to be the best kernel there is. No simple monolithic approach then, taken by a young guy who has just recently finished reading Tanenbaum’s book.
And BSD, if BSD hadnt been in a lawsuit then, Linux couldnt have been known. Now there was a empty time slot, where Linux jumped in and took it.
Edited 2007-08-11 12:25
I believe that they already had the furniture, the walls and the ceiling, just needed^aEUR| the rest.
You saying a kernel for an OS is like furniture for a built-up house ? Utterly stupid. What are you doing with all your fancy tools if you don’t even have a way to run them ?
I don’t see anything negative in that, despite the intent.
Me neither. Timing is everything. And Linux just happen to have it, at the time.
RMS and the GNU guys worked for decades trying to make a free os. they got the userland stuff done in a few years, but after about 8 or 9 years of effort, Hurd still wasn’t there. Linus on the other hand was attracting a huge group of enthusiasts to his kernel project, and quickly ended up with a usable kernel, but no userland. Linus chose GNU as one of a few options available to him (BSD userland being another). It isnt like the GNU stuff was rocket science or anything, these are all apps that had been re done over and over through the years.
That is what he is saying NOW, after almost 20 years of development and no usable kernel in sight.
Basically, what pisses RMS off is that he codes for philosophical purposes, he doesn’t want people to use his product because its good, he wants people to use it because he believes it is the only ethically right choice to make. Then this Linus guy shows up who doesn’t give a damn about his philosophy, but wants to write code because he finds it cool. The Linus approach catches on like wildfire, it turns out that the “lets all work together and make something awesome” approach works a hell of alot better then “software has freedoms just like people” approach when it comes to attracting developers. Contrary to popular belief, THAT is the greatest contribution Linus ever made to the open source world, showing people how much hobbyists working together can accomplish. The defining moment of open source is when ESR wrote an anthropological paper on what Linus stated called “The Cathedral and the Bazaar”, codifying the principals that made the linux project work so well, becomming what is known now as the open source methodology. Virtually overnight, people who were previously writing “Free” software, began writing “open source” software, and the free software movement again became an obscure band of ragtag campus revolutionaries.
That is why RMS is pissed at the whole Linux thing. He always says Linus is “just an engineer”, while Linus has no time for philosophical sillyness when it comes to software, and does what he does for practical reasons alone.
“What would you think if you built a house from the ground up, and suddenly someone comes in and puts furnitures there and gets the credit?”
Actually, it’s more like RMS had built furniture and other components but was lacking a foundation for the building. Linux created the foundation.
“And Linus doesnt want to give Stallman enough credit too. ”
Creadit for what? RMS/FSF had very little, if any, to do with the Linux kernel and that’s what Linus created; the kernel. He didnt create the distros, which perhaps should be called GNU/Linux distributions.
I use PCLinuxOS because it is free,it does not control
what hardware i have installed,it installs in 20 mins,
it is very very easy to use,it can play and copy ALL
dvds,it streams all major radio stations,it has a repository of thousands of programs that install in a
minute and always work,and lastly the most exhillarating
reason of all, i can VISIT ANY WEBSITE AND DOWNLOAD
ANY FILE WITH IMPUNITY.
To all Windows users,have you run your Spybot Search &
Destroy today? Guess what, it doesn’t catch all the
trojans.
Why on earth would i use Windows unless i was a big time gamer with no life!