Thinking Small

Wednesday, January 07 2004 @ 01:22 AM EST

Contributed by: PJ

Richard P. Feynman, in his 1959 speech "There's Plenty of Room at the Bottom - An Invitation to Enter a New Field of Physics," said that "in every theorem, there are assumptions".

But assumptions are not always accurate, which is one good reason for the scientific method of sharing ideas, of course. SCO, once again, appears to be suffering from a legal theory based upon flawed assumptions. I gather this from a comment left on Groklaw earlier today.

Here is the comment, from Trevor G. Marchall, Contributing Editor, (who added "BYTE is not associated with this post") and note that I don't know if it really is from him or not but for the purposes of our theoretical discussion, it doesn't actually matter:

"When I interviewed Mr Stowell he mentioned Linus

"I interviewed Mr Blake Stowell on 13 June 2003. I recall him specifically mentioning that Linus' university had a Unix Source license. Helsinki, if I recall correctly... I recall him speaking to this very issue - that Linus had been tainted by his University's Unix source license.

"Trevor G Marshall
Contributing Editor, (BYTE is not associated with this post)"

This brought this question:

"What do you mean by 'tainted' ?????????"

Marshall clarified with two responses:

"I recall this part of the discussion centered on students at universities, worldwide, who have seen Unix source code at their universities, and may have used some of that 'IP' in later projects, so that those projects might therefore be 'tainted' by the 'IP'

"Now I am not saying I agree with this line of argument, I am reporting my interpretation of what is in the notes I took during the discussion with Mr Chris Sontag


"Actually, I just noticed my notes say that Mr Chris Sontag raised this point, note Mr Stowell.


The exchange inspired pfusco to become creative and explain what SCO means by "tainted":

"I believe that he is referring to something that a SCO employee would say. Tainted... as in Linus saw the Unix code while in university and for some strange and bizaar reason it stayed in his head and cried out during the night 'put me in linuxxxxxx.... put meeeeeeeeee in Linux....' over and over again untill one day Linus succumbed and inserted the code."

This burst of creativity elicited from a rational soul the point that tainting is generally a trade secret or NDA concept:

"Interesting, but only if Linus saw the code and agreed to an NDA

"Linus never saw such code to my knowledge. And even if he did they can't say he's in breech of contract if he never agreed to the license, which I don't believe he did. That would leave them with trying to claim that his knowledge obtained from reading the Unix source (assuming he did) somehow leaked into the Linux source. 'Tainting' is usually something mentioned with trade secrets and non-compete agreements where it is significantly uglier and more difficult to avoid liability when working on two similar products."

This brought out the cynic in PolR, who pointed out how nonsensical it is to think of trade secrets being taught in university classrooms:

"Yeah! Trade secrets are taught in university classrooms.

"See the title. You think they can expect students to study and never use what they are taught? Would a judge really believe the student is at fault if he uses what he learns?"

Now, I haven't seen Trevor's notes, so I am just putting forth my own theory, using assumptions of my own, which may be as flawed as SCO's.

Well. Not *that* flawed, probably.

Let me be sequentially creative, cynical, and then rational. Let's say, putting on my creative hat first, that SCO's theorem is that all students who took a class in Unix are now tainted. Anything they do is now a derivative work of Unix, because the students saw the precious (and highly viral, I must say) Unix code in a class at the University of Helsinki or MIT or NYU back in the 80s. By this ka-ching, GPL-yang theory of SCO's, combined with their creative definition of derivative works, every operating system in the world is now their derivative work, is it not?

Now, I put on my cynical hat and ask myself, what would the consequences be? -- The whole world owes them money.

Silly, you say? Impossible? Not if you think big, and we have seen that whatever SCO's faults -- and they are legion -- thinking big is not a SCO problem. Note this BYTE article by Marshall from June.

Let's assume that this is the plan, long enough to see if it could float. My rational hat, please. My first rational question must be: did Linus study Unix in school as alleged? How to know? Well, I could ask him. So, Rational Me did, and here is his response when I sent him Trevor's comments:

"I can pretty much guarantee that that isn't the case. The University of Helsinki was a big VAX/VMS user, and as far as I know only started using unix after I had already started there - I started in 87, I think (with a year off in the army, so my second year would have been 89-90).

"And I believe that the 'UNIX and C' course I took in that second year (might even have been the third year, I forget) at university was the first time they ever taught Unix. And that was an introductory course in _using_ UNIX and C, not in internals.

"And they used the Andrew Tanenbaum book in the Operating System class, which again has no UNIX internals, it uses Minix as an example.

"So I not only never saw any UNIX sources lying around - I'd be very _very_ surprised if the University of Helsinki ever had a source license. It just wasn't their thing. The university didn't do any OS research at that time (Linux kind of re-stimulated some of it, and they have since done some stuff), and hadn't done OS research since the 60's, I think. That's how they got into Burroughs and then later VAX/VMS.

"Sounds like our dear SCO people are just making things up again."

Well, that is if Trevor is who he says he is and his notes are accurate, etc. Assumptions. If all those assumptions are accurate, then SCO's problem, or one of them, would be that while they posit creative theorem, they don't bother to check them against reality before running with them. If Linus never saw Unix at school, and he says he didn't, their concept goes poof. Although everything else in this article may be assumptions, this is a solid fact. There goes that legal theory.

You can't just figure a thing must have happened, and then go sue somebody, using discovery to find out if it did or did not happen. Not liking the scientific method, which is based on sharing information, can definitely lead to black hole dropouts in your reality base. That can gum up your legal works, no doubt about it.

Speaking of reality, more or less, you might enjoy looking at something one of the lawyers I do work for sent me tonight. I was, appropriately enough, in the middle of reading Feynman's "think small" speech when I got his email. How would you like to look at the universe smaller and smaller, from the Milky Way down to quarks on a leaf? If you like such things, or you just need a break from SCO's small and crass universe, go here and enjoy the view while traveling from the Milky Way Galaxy viewed from a distance of 10 million light years and then zooming in toward Earth in powers of ten to finally reach the leaf, and then zoom in to the level of the quarks viewed at 100 attometers. That's assuming I understood it and am accurately describing it.

Sometimes when SCO is getting to me, it helps to think about infinity -- how magnificently infinite the Universe is, both macroscopically and at the microscopic level. Which is another way of saying that tonight, my antidote to small thinking was to think exquisitely small.