Memory management and patents

[Posted August 14, 2002 by corbet]

Linux VM hackers are engaged in ongoing discussions on both large page support (covered last week) and improving the performance of the new reverse mapping mechanism. That conversation slowed down, however, when Alan Cox pointed out that a number of the techniques being discussed are covered by SGI patents. In fact, a closer look by Daniel Phillips shows that a number of existing Linux technologies, including reverse mapping in general and the buddy allocator, are covered by these patents. This is a problem, he said, that we can't ignore.

That was Linus's cue to jump in with his policy on software patents and kernel code:

I do not look up any patents on _principle_, because (a) it's a horrible waste of time and (b) I don't want to know.

The fact is, technical people are better off not looking at patents. If you don't know what they cover and where they are, you won't be knowingly infringing on them. If somebody sues you, you change the algorithm or you just hire a hit-man to whack the stupid git.

Linus followed up with a note that the above "may not be legally tenable advice." But he sticks by his point that, anymore, it's impossible to write an interesting program without running into somebody's patent. Rather than worry about it, it's better to just proceed and deal with any problems as they emerge.

This is probably the only rational approach; otherwise kernel hackers would go nuts trying to find and avoid all of the applicable patents. It's probably only a matter of time, though, until one of these patents bites the kernel in a big way - at least in the U.S. Those are the times we live in, though.

(Log in to post comments)

Memory management and patents

Posted Aug 15, 2002 7:48 UTC (Thu) by leandro (guest, #1460) [Link]

It would be very interesting if patents turned the Linux kernel – or even the whole GNU system – horribly expensive and virtually proprietary in the US, and perhaps NAFTA and EU areas too, while still dirty cheap and free in South America, Asia and Africa.

That could help complete the takeover of the First World by the Third World, which has already been supplying the population deficit of the hedonistic West. Now we would supply technology too.

Are the barbarians at the gates yet?

Memory management and patents

Posted Aug 15, 2002 8:16 UTC (Thu) by drooypcom (guest, #3272) [Link]

"That could help complete the takeover of the First World by the Third World, which has already been supplying the population deficit of the hedonistic West. Now we would supply technology too."

Actually I dont think this would work, at least for the third world. Because sure you could develop and use the Free Linux in the third world, but you could not sell it. In the third world because nobody has any money, in the first world because of the patents.

But! This could work for the "second world", in which I include good part of Asia, because there you can sell...

That the same old story: the middle class can take over the upper class, become the upper class and some times later the former upper class now middle class can take over the upper class again, in an endless cycle (see Orwell's 1984) but the lower class, the really poor peoples, they cant do that (as a class, but there is still hope for individuals to climb the socila ladder)

Memory management and patents

Posted Aug 16, 2002 17:13 UTC (Fri) by cekim (guest, #3295) [Link]

Rather than an ongoing class struggle, we are more likely to enter an
IP cold war built on the principle of Mutually Assured Destruction:

"I won't sue you over your infringement of my red patent if you won't
sue me over my infringement of your blue patent."

Alas, the Green patent was GPL'd so no one gets to color TV...

I am all for IP in principle, but something has to give in the US as
our Patent Office appears to be either incompetent or intentionally
disruptive to innovation...

I am expecting a patent on "Human Oxygen Intake" any day now...

Memory management and patents

Posted Aug 12, 2003 1:42 UTC (Tue) by leandro (guest, #1460) [Link]

> I am all for IP in principle

So you are all for a propaganda device? Intellectual property doesn't exist... neither copyrights nor patents are property, but a state concession of an artificial monopoly.

Memory management and patents

Posted Oct 4, 2002 10:32 UTC (Fri) by leandro (guest, #1460) [Link]

> you could develop and use the Free Linux in the third world, but you could not sell it

But that is the point. It is free software, no need to sell. You just operate it, do services, learn, whatever. Take a scrap [34]86 and run you backyard business on it.

Memory management and patents

Posted Aug 12, 2003 1:45 UTC (Tue) by leandro (guest, #1460) [Link]

> some times later the former upper class now middle class can take over the upper class again, in an endless cycle (see Orwell's 1984)

What has 1984 to do with this cycle? It is precisely about a distopic future where society is fossilized by "the party"

Memory management and patents

Posted Aug 23, 2002 22:11 UTC (Fri) by josmala (guest, #3401) [Link]

IF anyone in hardware industry would DARE to sue linux kernel developers for their software patents the result would probably hurt them a LOT more than if they would of simply given up on their patents.
Consider if your company would be considered the great evil by linux developers, AND amound of bad PRESS they would definitely get, there would probably be EVERY possible bad aspect of the company coming publicly in technical world. Their sales figures would definitely drop within few months by figure by loosing their image. The stain that they would get in their image would be such that no amount of marketing money could fix it.
And if the company it self would use linux they definitely wouldn't do it since they might loose some valued employees to other companies, and hiring when next upturn comes would become harder.
If the suing company would be like M$ they would probably get extra monopoly investigations and it would be a risk for them too.
What about other software companies. Well if they HOPE to sell their products in Linux market some time they had damaged their reputation and that hurts. Also they might find them self in trouble with their current customers. And suddenly have harder time to find good programmers.
To simply put NO ONE is goint to use their patents for linux kernel, they got too much LOOSE if they would even try it.
Most wouldn't consider much of a damage if linux kernel would use the algorithm they have patented.
Only reason they could use that would be that someone in linux industry would sue them over patent, and then they would deny the other company for using kernel versions that use their patents. [Consider if redhat would sue SGI with some software patent, SGI could tell redhat NOT TO sell/use kernel version 2.6 which would incorporate their patents, as a counter suit, or more easily, show RH that they could sue them with that, and let them decide weather stop shipping any products that use the patents or withdraw their case, easy choise don't you think, SGI could use the patent just to settle a court case with no one outside case to ever know what really happened.]

Memory management and patents

Posted Aug 24, 2002 14:42 UTC (Sat) by boebert (guest, #3412) [Link]

I strongly suspect the bulk of these patents can be trivially invalidated by citing parts of the Multics design as prior art. It makes a strong case because it was both implemented and (for its day) fairly widely deployed. I would put up a plea for assistance on alt.os.multics. I wouldn't waste a lot of time though -- some of us are rapidly approaching the drool-cup stage :-)

Cheers

Earl