Man AdTI Hired to Compare Minix/Linux Found No Copied Code (SCO vs. IBM/Linux thread)

Groklaw ^ | Thursday, May 27 2004 @ 05:01 PM EDT | Pamela Jones

[shadowman99]

Man AdTI Hired to Compare Minix/Linux Found No Copied Code
Thursday, May 27 2004 @ 05:01 PM EDT
Andrew Tanenbaum has published the most remarkable email from the man hired by Ken Brown to do a line-by-line comparison of Minix and Linux, Alexey Toptygin, who summarizes his findings and posts them on the Internet: "Around the middle of April, I was contacted by a friend of mine who asked me if I wanted to do some code analysis on a consultancy basis for his boss, Ken Brown. I ended up doing about 10 hours of work, comparing early versions of Linux and Minix, looking for copied code. My results are here. To summarize, my analysis found no evidence whatsoever that any code was copied one way or the other." When he turned in his work, he had a conversation with Brown: "Apparently, Ken was expecting me to find gobs of copied source code. He spent most of the conversation trying to convince me that I must have made a mistake, since it was clearly impossible for one person to write an OS and 'code theft' had to have occured. So, I guess what I want to say is, pay no attention to this man. . . " Eric Raymond has also answered Ken Brown's Samizdat. Another very detailed response here, on Newsforge, by Jem Matzan. I'll end your suspense. No, they didn't like it. Matzan: "In the history of publishing there has never been a less scrupulous work than this book. It's a stinging insult to real books and genuine authors everywhere, harming the credibility of all of us who write for a living." Raymond publishes his email to AdTI, who inexplicably (unless the book is an elaborate troll) and foolishly sent him a copy to review: "Judging by these excerpts, this book is a disaster. Many of the claimed facts are bogus, the logic is shoddy, some of the people you claim to have used as important sources have already blasted you for inaccuracy, and at the end of the day you will have earned nothing but ridicule for it. . . . "The problems start in the abstract. Software is not composed of interchangeable parts that can be hodded from one project to another like a load of bricks. Context and interfaces are everything; unless it has been packaged into a library specifically intended to move, moving software between projects is more like an organ transplant, with utmost care needed to resect vessels and nerves. The kind of massive theft you are implying is not just contingently rare, it is necessarily rare because it is next to impossible. . . . "Your account of the legal disclosure history of the Unix source code is seriously wrong. Persons authorized by AT&T did, in fact, frequently ship source tapes which contained no copyright notices — I know, because I still have some of that source code. . . . "I began reading the excerpts skeptical of the widespread conspiracy theory that this book is a paid hatchet job commissioned by Microsoft. Now I find this theory much more credible. I can't imagine how anyone would want their names on a disgrace like this unless they were getting paid extremely well for undergoing the humiliation. . . . "You claim that 'To date no other product comes to life in this way', presenting Linux as a unique event that requires exceptional explanations. This is wrong. Many other open-source projects of the order of complexity of the early Linux kernel predated it; the BSD Unixes, for example, or the Emacs editor. Torvalds was operating within an established tradition with well-developed expectations. "'Is it possible that building a Unix operating system really only takes a few months —and, oh by the way, you don't even need the source code to do it?' Yes, it is possible, because there are published interface standards. I might have done it myself if it had occurred to me to try — in fact, I have sometimes wondered why it didn't occur to me. "As for whether it was possible to produce Linux in the amount of time involved — it is never wise to assume that genius programmers cannot do something because the incompetent or mediocre cannot. Especially when, as in Linus's case, the genius already has a clear interface description and a mental model of what he needs to accomplish. . . . "You propose that the absence of credits to developing countries might be evidence of some sinister memory-hole effect. The true explanation is much simpler: developing countries don't have Internet. There is a straight-up geographical correlation between contributions to open-source projects and Internet penetration." There is a great deal more, and I encourage you to visit all four sites, to get the complete picture. Honestly, how incompetent must you be to think attacking Linus Torvalds' integrity is a good strategy? He is loved and admired internationally by folks who do understand the code, unlike Mr. Brown, and everyone knows such a man would never knowlingly steal anyone's code, period. Nobody else would either. It's not the FOSS way.

This work is licensed under a Creative Commons License.

[antiRepublicrat]

You're actually trying to sell the BS that Torvalds had no choice but name the code after himself?!?

You know nothing of social dynamics. Once the name was started by his friend, and everybody hated his name, Linux stuck.

Not remarkably, his kernel incorporates concepts from this source code, such as the filesystem, source tree, etc.

Remarkably, Linux is so vastly different in internal structure from MINIX that Tanenbaum was its first critic. Yeah, he must have copied everything. It's like sitting down with Mercedes blueprints and coming out with a Lexus.

You finally admit that Torvalds isn't the inventor of Linux.

And Ken Thompson wasn't the inventor of UNIX because MULTICS already exsisted. BS. I guess Thompson was a MULTICS parasite. And I guess you'll never say Bill Gates was a parasite of almost everyone.

I've been very clear about saying that he wrote the code. See #8. Invent another strawman.

Yes, you were very clear and toned down your claims of theft AFTER you were called on it.

Torvalds consulted the MINIX book and had a significant amount of kernel source code at his disposal.

It's funny, you portray Tanenbaum as the poor man whose original work was stolen for Linux, but you don't criticize Tanenbaum for the fact that his OS was just a copy of UNIX.

Along your lines of thinking, NOBODY every invented any OS, they all just stole from the previous. But you won't criticize anyone but Linus for their OS having some sort of ancestry, even those who exactly copied code.

[Bush2000]

Hmmm. That so? You're sure about that? Remind me again who got the Windows source code for free?

That's hilarious. You OSS guys want to equate Microsoft's allowing the Chinese to sit in a room with a bunch of lawyers to view the Windows source code under controlled circumstances -- with you and your Eurotrash OSS buddies toiling away -- fat, dumb, and happy -- working on source code and updates for the ChiComs as slaves. Hilarious. Do you do stand-up, by any chance?

[Bush2000]

You know nothing of social dynamics.

More spin. Torvalds supposedly wrote the Linux kernel. He could have named it anything he wanted. It's baffling why you guys can't even admit that simple fact.

Yeah, he must have copied everything. It's like sitting down with Mercedes blueprints and coming out with a Lexus.

Your posts are becoming increasingly erratic. I never said he "copied everything". Read for comprehension. But there are certainly large portions of code that borrowed liberally from MINIX.

And Ken Thompson wasn't the inventor of UNIX because MULTICS already exsisted. BS. I guess Thompson was a MULTICS parasite. And I guess you'll never say Bill Gates was a parasite of almost everyone.

Anybody who leeches onto somebody else's work and calls it their own is a parasite.

Yes, you were very clear and toned down your claims of theft AFTER you were called on it.

You're a liar. I never claimed he outright copied code. That is your strawman. Beat your dead horse on your own time.

It's funny, you portray Tanenbaum as the poor man whose original work was stolen for Linux, but you don't criticize Tanenbaum for the fact that his OS was just a copy of UNIX.

Tanenbaum ain't the issue. Stop changing the subject.

Along your lines of thinking, NOBODY every invented any OS, they all just stole from the previous. But you won't criticize anyone but Linus for their OS having some sort of ancestry, even those who exactly copied code.

Again, you're simply trying to change the subject. Why do you OSS guys have such short attention spans -- A.D.D.?

[Bush2000]

No, the point of this thread, if you would RTFA, is that Ken Brown hired a guy to find where Linus copied Minux code, and when the guy couldn't find such code, Brown carried on anyways.

Notice I said "Brown's article", not "this thread". I realize that reading must be difficult for you, but give it a try every once in a while.

[Bush2000]

Programming is an art form.

Too bad you can't patent art.

[antiRepublicrat]

Torvalds supposedly wrote the Linux kernel. He could have named it anything he wanted.

And for the public release he named it Freax. It didn't stick.

But there are certainly large portions of code that borrowed liberally from MINIX.

Borrowed = copied. The MINIX author and Brown's own researcher says that didn't happen. Get that into your head!

Anybody who leeches onto somebody else's work and calls it their own is a parasite.

Everybody remember this day! Bush, our resident Microsoft astroturfer, just called Bill Gates a parasite. Careful Bush, you're going to lose your job.

I never claimed he outright copied code.

Hey everybody, Bush walked up to Linus with a gun and squeezed the trigger. Linus is now dead! Don't try to claim libel. After all, I didn't actually say you killed him.

[Bush2000]

And for the public release he named it Freax. It didn't stick.

Poor, helpless Linus. He lacked the ability to name his own pile of bits. /SARCASM

Borrowed = copied. The MINIX author and Brown's own researcher says that didn't happen. Get that into your head!

You can't seem to get this through your thick cranium. Borrowed means he sucked up the concepts and adapted them for his own use in Linux. It doesn't mean he copied the code:

Tanenbaum: "I told him that MINIX had clearly had a huge influence on Linux in many ways, from the layout of the file system to the names in the source tree, but I didn't think Linus had used any of my code."

Hey everybody, Bush walked up to Linus with a gun and squeezed the trigger. Linus is now dead! Don't try to claim libel. After all, I didn't actually say you killed him.

Dude, you're unstable. Seek counseling.

[TechJunkYard]

Too bad you can't patent art.

Precisely.

[antiRepublicrat]

You can't seem to get this through your thick cranium. Borrowed means he sucked up the concepts and adapted them for his own use in Linux. It doesn't mean he copied the code:

"Borrowed" concepts, fine. But you said "borrowed code" which means copied, stole. Keep up the backtracking Bush. That plus if you'd read the article, you'd know that Brown assumed there would be copied code in Linux and counted on it for support of his book. Copied code was to be the centerpiece of his argument. Notice how he went into denial when he found out there was none.

But minus the support for the central theme in his book he published anyway. Really honest, huh?

Dude, you're unstable. Seek counseling.

You're trying to say that because you never used the word "stole" you didn't claim he stole. But you've used every other word in the book to say the same thing.

[MarkL]

I never said that Torvalds physically copied MINIX code. But he most certainly took the ideas directly from MINIX and incorporated them into his code. He didn't invent the ideas. He stood on the shoulders of giants, and then had the audacity to name Linux after himself.

Hmmmm... I doubt that you'll fail to find some influences on any former computer science student who's now an OS system programmer by Andy Tannenbaum... Or compiler writer, for that matter, since both of his books on the subjects are pretty much considered to be "standards" in most CS curriculems (sp?). BTW, building on the ideas of others IS pretty much standard in the programming world... Otherwise, we'd still be using paper tape, and there'd be no such thing as virtual memory. Don't forget, UNIX was built on top of ideas gained from the MULTICS projects in the first place. But just because someone builds upon, or improves upon an idea of another does NOT mean that theft is involved. Afterall the mouse was "invented" at PARC, but the design of the optical "3d" mouse isn't called theft.

Mark

[Bush2000]

I don't fault Torvalds for doing what he did; after all, building on the ideas of others is the primary way that our civilization advances. I should add that he did a pretty good job, given his lack of experience and the primitive tools at his disposal. Regardless, he achieved something that isn't very common -- and that should earn him the admiration of many for doing so.

But ... I object to the attitude of many Linux advocates here who believe that Torvalds wrote his kernel in a vacuum -- without any influences and without the help of a reference model such as MINIX. I don't know why they have such a difficult time admitting the obvious. I believe that they think that admitting it will cheapen Linux in some way; whereas, in their minds, they've imparted some kind of god-like status on Torvalds which isn't fitting.

But ideology dies hard.

[Bush2000]

"Borrowed" concepts, fine. But you said "borrowed code" which means copied, stole.

I'd like you to show me where I said "borrowed code" [hint: I didn't].

But minus the support for the central theme in his book he published anyway. Really honest, huh?

No, the central theme in his book is that Torvalds could not have written a kernel in 3-4 months without consulting some kind of reference model such as MINIX. And, given the evidence that I pointed out above -- that (1) Torvalds had Tannenbaum's book, (2) the book contained 12000 lines of code from the MINIX kernel, and (3) Tanenbaum found that Torvalds had borrowed many concepts from MINIX (filesystem, etc) -- Brown is correct.

You're trying to say that because you never used the word "stole" you didn't claim he stole. But you've used every other word in the book to say the same thing.

If you're upset because I didn't use the word "stole", that's your problem. I've been very clear in my posts.

[antiRepublicrat]

I'd like you to show me where I said "borrowed code" [hint: I didn't].

You: "There's simply no way that he generated that amount of code in that short of a time period without borrowing code from other sources." Oops.

No, the central theme in his book is that Torvalds could not have written a kernel in 3-4 months without consulting some kind of reference model such as MINIX.

Did you read the article?

Ken kept pestering me to hurry up and finish. He told me he had a paper awaiting publication, and that my analysis was the last bit of data he needed. ... my analysis found no evidence whatsoever that any code was copied one way or the other. ... Apparently, Ken was expecting me to find gobs of copied source code. He spent most of the conversation trying to convince me that I must have made a mistake, since it was clearly impossible for one person to write an OS and 'code theft' had to have occured.

Brown's book depended on there being copied code, and Brown was in denial when none was found. His idea wasn't that someone couldn't write an OS in three months without having a model to go on as you've changed your story to, but that it couldn't be done without stealing code because no one apparently could type in that much original code. The base of his book, and of your arguments, fell out from under him.

I've been very clear in my posts.

Yes, you've been clearly parroting Brown. Yet his own researcher, the supposedly wronged person and the world's foremost UNIX historian have all called his work complete BS.

[MarkL]

I don't fault Torvalds for doing what he did; after all, building on the ideas of others is the primary way that our civilization advances. I should add that he did a pretty good job, given his lack of experience and the primitive tools at his disposal. Regardless, he achieved something that isn't very common -- and that should earn him the admiration of many for doing so.

But ... I object to the attitude of many Linux advocates here who believe that Torvalds wrote his kernel in a vacuum -- without any influences and without the help of a reference model such as MINIX. I don't know why they have such a difficult time admitting the obvious. I believe that they think that admitting it will cheapen Linux in some way; whereas, in their minds, they've imparted some kind of god-like status on Torvalds which isn't fitting.

But ideology dies hard.

I don't think that it's so much ideology, as much as terminology. You're right, of course, that Linus didn't do the work he did in a vacuum, but at the same time, calling what he did an "adaptation" isn't quite right either. Here are the facts... 1) Linus DID write the code for the kernel. 2) Many consider the kernel to the the OS itself, i.e. Linux. 3) As he did NOT use existing code and modify it (which would be an example of an adaptation), calling what he did an "adaptation" isn't what he did. OTOH, if you use a "literary" definition of "adaptation," then yes, Linus adapted the ideas and concepts to Linux.

Linus came up with a "work-alike" OS. This really isn't that different than another *IX work-alike (sort of) known as MS-DOS.

The key is that he took classes that informed him as to what makes up a "good" OS. He then wrote a kernel that took advantage of those features. But remember, that the kernel is NOT the same as "Linux," however, for the collection of software that's known as "Linux" to really BE Linux, it needs that kernel. And also remember that the kernel has interfaces to things like libraries and file-systems. And there's no such thing as an OS that magicly springs out of the ground. As we saw (and you state) the original version of the Linux kernel was distributed in source code format. Linus publicly stated that Minix was needed to compile the code. And it depended on the Minix file systems and other support. But eventually, Linus and many others added Linux specific features, allowing the actual Minix features to be eliminated. Does that mean that they didn't use the best features of things like the file system? Hell no! If you've got something that you know works, why start from scratch? Did David Cutler and friends start from scratch with NTFS? No! They simply improved upon HPFS.

So there is some reason to say that Linus "invented" Linux. Without the Linux kernel, there is no Linux. On the other hand, since AIX was based on the Carnagie-Mellon MACH kernel, using that same sort of reasoning, one could say that CM "invented" AIX. And that doesn't quite fit either.

Again, the problem here is simply one of terminology.

Mark

[ShadowAce]

I object to the attitude of many Linux advocates here who believe that Torvalds wrote his kernel in a vacuum -- without any influences and without the help of a reference model such as MINIX. I don't know why they have such a difficult time admitting the obvious.

Who are "they?" I'm really curious, since I've seen no posts by anyone here stating that, or even implying it.

[myself6]

Even if true. It is not illegal, or against copyright laws. Its like saying that I cannot read a book to learn how to write.

[antiRepublicrat]

I thought you were trying to split hairs over where he borrowed ideas from. MINIX, Posix,

Copying from Posix. You're getting more entertaining as you go.

[antiRepublicrat]

Cars, Ford. Good point, bad example. How about Lexus? We knew Lexus was going for the Mercedes market when the brand started, and they knew from Mercedes exactly what constitutes an excellent car in that market segment. Yet Lexus went on to make excellent luxury cars in their own right. Mercedes copy? No way since we haven't seen a torrent of lawsuits. Mercedes inspired? Of course. And Linus doesn't deny his inspiration.

[Havoc]

You might try reading some interviews or, forbid, watch the movie "Revolution OS". Many interviews with Torvalds, Stallman, etc. Gets into linux and GNU developed. And Linus addresses how the code came about on the record. I'm sure you'll be surprised - even shocked and pasty white when you learn that Linus did in fact develope the OS on his own. You guys are begging for a fallback position now that allows you to try to diminish Linus but that doesn't make you look quite as bad. Hint - damage is done. Learn, correct your mistakes and get some respect. No harm no foul.

[Havoc]

Thank you Shadowman! That brings back some memories. Just add the GNU project, put on lid, shake vigorously and pour out MS kryptonite in healthy portions LOL.