First Analyst Impressed By SCO's 'Proof'

informationweek.com ^ | June 5, 2003 | John Foley

[NotQuiteCricket]

One of the first observers to see the evidence in the SCO Group Inc.'s intellectual-property claim against IBM and the Linux community says there are direct similarities between the Unix code that SCO claims control over and the Linux operating system. "My impression is that [SCO's claim] is credible," says Laura DiDio, a Yankee Group analyst who was shown the evidence by SCO Group earlier this week. "It appears to be the same" code. But DiDio says the developing battle could hinge on legal fine points that are hard to sort out in the current atmosphere of claims, denials, and counterclaims.

Apparently the most telling evidence is that parts of the SCO code and Linux code include identical annotations made by developers when they wrote the programs, says DiDio, who compares such notes to the signature or fingerprint of a developer's work. "The fact that these appear to be transposed from Unix System V into Linux I find to be very damaging." DiDio says she was shown several instances where the source code and developer's comments in one operating system were the same as in the other operating system.

[NotQuiteCricket]

Go do a search on Laura DiDio (Yankee Group analyst) and find out what she sells. Humn? MS has their fingers in this *again?* I was just curious who this analyst was and how much experience she had with linux, and apparently she is a MS partner. There is a big fat question I had which is : where did the Linux code come from? Did she download it and bring it to the meeting, or did SCO provide it? Cause if it wasn't fresh from the internet who is to say that SCO isn't lying? And lastly, why hasn't the information been released to the public yet?

[martin_fierro]

Ping

[tomakaze]

the SCO/LInux thing explained

[kevkrom]

I'm shocked, shocked, I tell you that a Microsoft-dependent research firm would produce a report by a non-programmer analyst that would make these claims.

</SARCASM>

[detsaoT]

If the SCO code segments in question appear anywhere within the Linux kernel, they win. Particularly if the Linux developer that ported the code over was so lazy, they didn't even remove the embedded comments in the code.

If you are afraid of this lawsuit, play it safe. Run FreeBSD. Certified UNIX-Free out of the AT&T/USL/Novell lawsuit of 1994. Heck, FreeBSD is more stable than Linux anyway - that should be reason enough.

(Ok, shameless plug. So sue me. ;) )

;) ttt

[Only1choice____Freedom]

If the SCO code segments in question appear anywhere within the Linux kernel, they win.

This isn't necessarily true. They have to prove exactly where it came from. If it comes from Linux to Unix then it is a problem for SCO.

[Ernest_at_the_Beach]

Hmm!

What about the Novell statements on their ownership?

OFFICIAL BUMP(TOPIC)LIST

[JSteff]

That site is excellent! SCO is a failing entity and should just go away. Novell still has all the rights to the source code anyway... at least from what I read.

[detsaoT]

This isn't necessarily true. They have to prove exactly where it came from. If it comes from Linux to Unix then it is a problem for SCO.

Yes, this is true. I do think that SCO would be able to prove that their codebase in question is older than the corresponding Linux code, or they wouldn't be bothering with a potentially fatal (for them, if they lose) lawsuit.

Linux source code is in the public domain. It should be a trivial matter to determine at what point in time the code in question was imported into the kernel (and, if SCO is suing on the premise that "Linux" is operating on copyrighted code, then the kernel would be the only component of "Linux" that could possibly be affected, as the add-on tools are not "Linux"). If SCO can prove that the source existed in earlier releases of their codebase (in either OPENserver or UNIXware, or the predecessor products), then they win.

This is actually a potentially devastating case for Linux. Us BSD'ers went through the same thing with SCO's predecessors, and the resulting lawsuit actually set back our projects by a few years. Even if SCO wins, they won't be able to really punish anyone other than potentially IBM (I would assume that they are accusing IBM of porting their "proprietary" code over to Linux) over it - as a result, someone (most likely Linus, since He Owns The Kernel) will purge the offending code from all of the CVS branches, and the resulting Kernel will be rubber-stamp approved for future development.

I will be sorely pissed if this "proprietary" technology is something simple. This had better be a major subsystem in the kernel, SCO.

:) ttt

[AdA$tra]

Linux = Unix for all practical purposes. SCO has a good case. Open source movement is going to take a hit.

[HiTech RedNeck]

So why isn't this still a Novell issue, not an SCO issue?

[detsaoT]

Linux = Unix for all practical purposes.

Actually, Linux is less UNIX than the Berkeley codebase. BSD is over 20 years old, and the initial development was tightly intertwined with System V code (the BSD code actually introduced the TCP/IP network stack which brought about the modern internet, which was imported later into SYSV). Linux is actually only a little over 13 years old, and was purportedly written from scratch, as a UNIX lookalike, just like MINIX (the educational operating system that Linux was based on) was.

Linux != UNIX.

:) ttt

[eno_]

"Yes, this is true. I do think that SCO would be able to prove that their codebase in question is older than the corresponding Linux code, or they wouldn't be bothering with a potentially fatal (for them, if they lose) lawsuit."

It will be take more to prove this conclusively. UNIX kernels are, of course, older than Linux, but they have had enough development alongside each other that you would have to go through a source code control database for the particular development branch of the particular UNIX kernel in question and show that some particular engineer authored the code in UNIX first.

If they do it right, instead of just comparing two pieces of code, they should be able to say exactly who copied what from who.

SCO has already cast the die. If the lawsuit fails, SCO is very very burnt toast.

[thatsnotnice]

Heck, FreeBSD is more stable than Linux anyway - that should be reason enough.

They used to have a highly tweaked FBSD ftp server at cdrom.com that set some pretty impressive records for amount of data transferred at one time.

I haven’t even thought of them in a long time. I sort of stopped messing with FreeBSD around the time that Hubbard bailed out to work at Wind River or somewhere…

[Only1choice____Freedom]

SCO is just going after IBM's AIX code now. Linux is next. SCO must prove:
1. The code is the same.
2. The code came from SCO to IBM AFTER the AIX license was given.
3. The code was not in any other legal transfer.
4. They actually do own the code and not Novell.

If any of these conditions (maybe more) are not met, they cannot go further.

Even if IBM is in violation of sharing the code to AIX, they would hve to prove after that that it made it's way into Linux by way of IBM. I do believe IBM has been taking precautions. I also believe Linus has as well.

We'll find out soon enough. In the mean time, SCOX will jump up and down. The potential for making a quick buck on these guys is hugh.

[Only1choice____Freedom]

(Ok, shameless plug. So sue me. ;) )

Be careful what you ask for...

[HiTech RedNeck]

The potential for making a quick buck on these guys is hugh.

Not to mention series.

[Steven W.]

that is not exactly correct. SCO must prove beyond a perponderance of evidence is correct and then the burden will shift to Linux folks to prove it's not. In the case of SCO they'll probably be able to overcome that hurdle but there is absolutely no possible way, ever, that Linux folks can disprove anything in any way because of the nature of open source (in other words there is nobody available to testify fully & absolutely on Linux's behalf because there's nobody in such a position to do so)

[NotQuiteCricket]

"So why isn't this still a Novell issue, not an SCO issue? "

I have no idea - except that Novell is interested in linux contuning to live - (IIRC the article about the letter Novell put out had references to deals Novell was going to do with someone on the Linux platform). If all you do is read the articles on internetweek.com - there is no logic to the process that is going on in public - except possibly the pressure that MS is trying to put on linux through SCO. Novell's letter was a surprise to MS I think, and something that SCO tried to hide (the reference in the letter from Novell about several requests from SCO for Novell to sell the rights to the Unix patents/copyright). If the changes were made to Unix while SCO was developing Unix, then I guess they could have a case, but if the code (in Unix and/or Linux) pre-dates SCOs purchase by Caldera I'd guess that the case would then be moot (I think). While it is not amusing on an economic / open source level - it is funny that these corporations are acting like a pack of 7 year olds (or is that an insult to 7 year olds?).

Why didn't this article mention that the "analyst" works for MS / or is economically involved with/in MS products? Wouldn't that make someone think twice about her statement in the last paragraph?

[HiTech RedNeck]

Novell didn't give SCO an exclusive license did it? So it shouldn't be any skin off of SCO's nose, unless SCO can show that somehow Novell misrepresented how exclusively Novell intended to license out UNIX.