Torvalds: Linux not becoming obsolete would be 'sad'
I admire Linus Torvalds' candor (this is the guy who freely admits his own family doesn't use the Linux desktop, after all), as well as his foresight.
In an excellent interview posted on Simple-Talk, Torvalds covers a range of topics, including the Linux operating system's place in history:
I can certainly imagine the Linux kernel becoming obsolete--anything else would just be sad, really, in the big picture.
He is absolutely right. Much as we may pine for this or that project to achieve market dominance, it is one of the cardinal virtues of open source that there are no legal or business policies that would entrench it as a monopoly. People may choose to use it for a long period of time and to the exclusion of other products/projects, but there are no nefarious designs in the code to make it so.
Hence, Linux may fade away. At some point, we should certainly hope so, in order to make way for the next phase of operating system, one that is preferably open source.
Other interesting tidbits from the interview are Linus' comment on patents:
...[P]atents are very much used to stop competition, which is undeniably the most powerful way to encourage innovation. Anybody who argues for patents is basically arguing against open markets and competition...
And on working for Microsoft:
I find it unlikely that MS would ever offer anything that I would consider relevant. Money? Hey, they have it, and I like it, but I obviously don't value it over everything else. And they are unlikely to offer the things I really value.
A level-headed, mature commentary on the future of Linux, the value of patents, and the likelihood of Torvalds going to Microsoft. No wonder we love this guy.
Matt Asay is general manager of the Americas and vice president of business development at Alfresco, and has nearly a decade of operational experience with commercial open source and regularly speaks and publishes on open-source business strategy. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure.
- Topics:
-
Industry news,
-
Microsoft,
-
People
- Tags:
-
Linus Torvalds,
-
Linux,
-
open source,
-
Microsoft,
-
patents
- Bookmark:
- Digg
- Del.icio.us







There are a few kinds of technological advancement that we can observe in our lives;
1. Incremental-linear. For example: DDR3 over DDR2
2. Incremental-nonlinear. Hybrid HDD over normal HDD
3. Breakthrough. For example: DDR RAM over SDRAM or SSD over HDD
Patents are useful for a very small portion of 2 and all of 3 but should not be applied to 1 at all because that is where patents become counter productive. At some point every creation should become public domain perhaps 10 years giving companies sufficient time to recoup their investments and make a profit but not be able to hide behind their product forever.
There are a few kinds of technological advancement that we can observe in our lives;
1. Incremental-linear. For example: DDR3 over DDR2
2. Incremental-nonlinear. Hybrid HDD over normal HDD
3. Breakthrough. For example: DDR RAM over SDRAM or SSD over HDD
Patents are useful for a very small portion of 2 and all of 3 but should not be applied to 1 at all because that is where patents become counter productive. At some point every creation should become public domain perhaps 10 years giving companies sufficient time to recoup their investments and make a profit but not be able to hide behind their product forever.
Torvalds knows that he'd never sit back and say, "gee, Linux is obsolete" and twiddle his thumbs. He'd do something about it, or one of the other community grains of sand would.
"Patents are very much used to stop competition, which is undeniably the most powerful way to encourage innovation."
Reread what he said. He said patent is "a powerful way to encourage innovation." I think what he meant from that is the exclusivity to market a product from the patent provides a powerful incentive to innovate. Otherwise people would copy each other's work and compete with each other.
Basically, the agreement was like this: you document your invention in sufficient detail that a person reasonably skilled in the art could reproduce the invention. For telling everyone how to make your invention, we will give you 20 years of exclusive access. After that time frame they can use your invention for themselves.
In the modern world, there are two problems with this. 1) The pace of change means that many things are obsolete within 20 years so there is no public good from sharing the 'secrets'. 2) Patents are supposed to be 'non-obvious'. Basically, you are supposed to be revealing something of value. However, 'one-click' shopping is not only obvious, but it has no 'secret' that needs to be revealed. Anyone who hears of on-click shopping could reproduce the capability without needing to read the patent.
A final note, drug patents may still need substantial timeframes because it takes so long for regulatory approval and human trials before they can hit the market.
Our culture has created too many "cash cow" legal mechanics that have been exploited by corporations.
As for Linux going obsolete. We can only hope everything we use today becomes obsolete for better ideas.
A person who becomes very successful and highly respected in one field starts to think that he is an authority on all fields.
By his statement Linus Torvalds obviously knows squat about the legal and economic theory of patents. I'll listen to his ideas about software and operating systems. His ideas on patents, I'll let that slide.
Suppose Linus had actually developed a unique and useful technology only to have a large company copy it leaving him penniless, would he feel the same about patents? I suspect not.
In the meantime, the open source movement continues to destroy shareholder value by undermining innovation in the hardware and operating system world. Commoditization of hardware may mean lower prices, but commodities, by definition, will not allow for differentiation. Differentiation is what creates value. Most of the arguments being made in favor of Linux involve saving money on hardware. This is preposterous since Moore's law already made the hardware market very price competitive. Open source has done little to drive down software costs. Microsoft, and Oracle continue to fleece their customers selling them overpriced, buggy products. Meanwhile companies like Sun have sustained most of the damage. Go try and buy a copy of Oracle Database Enterprise Edition for any machine and you'll find that it's cost will often exceed that of the hardware it's running on. That's true for both SPARC machines running Solaris or a Dell Intel machine running Oracle's flavor of Linux.
Linux is already obsolete in my mind. It was from day one. Bell Labs solved the problem of a portable operating system (one that can run on any processor) decades before. There was virtually NOTHING new in Linux other than a Marxist pricing model. It never addressed the REAL problem of software costs. AIX., Solaris and HP/UX were never high margin products. They were the icing on the cake that made you choose one brand of hardware over another.
Based on your argument, everyone is a copy of everyone's else. There is not one person in this world who is not a copy of what he was taught because knowledge comes from the information existed. What Linus did is organize these knowledge into something people find useful. Bell Labs did the same. It copied and use Assembly to create C.
Commodization is good because more people can enjoy the product at a lower price. I remembered an Operating System costing as much as $2,000. And the Operating System is junk too compare to today's standard. If the operating system didn't commodize, we would never move up the innovation chain to have Applications. And if it weren't for Applications commodization, we wouldn't move up the innovation chain to have great Software Service like google search, web based bank app, on demand video, social networking myspace, etc.
You can have a spontaneous, random idea that will most probably be had by someone else in a short time, and you get the same protection as someone that got some solution to a problem through decades of directed (and expensive) research at their own risk.
A better system would only protect the result of directed research performed in an open way. You would apply for patent protection for the solution to problem X at least one year in advance. You would then perform your research ,knowing that, if you develop a solution after one or more years of research, you would get protection for some specified time (five to ten years sounds reasonable). If someone else comes up with a solution to your proposed problem (which was published openly at the time of application) then unless they were doing directed research and had applied for protection earlier, or the solution was obvious and didn't merit protection.
This way, random ideas do not get protection, and they will continue to be had at the same rate. Directed research is promoted, and people can invest in research knowing that, if they get to something useful, they'll be able to profit from it.
Everyone wins.
-
by estammer
July 21, 2008 10:11 AM PDT
- My thoughts prior to reading your article are that Linus thought the Linux OS should become Obsolete.
-
Reply to this comment
View
reply
-
-
See all 38 Comments >>However, after reading your article and the original documentation associated with it, the meaning of what he said is totally different.
While your article title is eye catching and brings the reader in, it is totally misleading for those who just peruse the titles of articles and move on.