Chris Brown attended today's hearing in SCO v. IBM and sends this quick message:
Taken under advisement. Rochkind declaration allowed, but IBM has 10 days for
expert to rebut it. More when I get home.
I won't comment until we have more details. But I knew you'd want to know as fast as I knew. Here's the SCO motion to file the declaration, which IBM opposed or in the alternative asked for more time to respond to, which is evidently what happened.
Update: More from Chris:
I've just returned from the SCO v. IBM hearing.
The short of it is that IBM's motion on the 192 items has been taken
under advisement. The court will accept the declaration of Marc
Rochkind, but will allow IBM 10 business days to submit an expert rebuttal.
David Marriott argued for IBM and Stuart Singer argued for SCO.
Before I write my full report, I'll mention a couple of quotes.
Judge Wells to SCO's Stuart Singer: "Does SCO have, can they provide,
additional specificity... I mean, basically, Is this all you've got?"
Stuart Singer replies: "There might be that in the last two months, that
more has been discovered, but yes, at the time of the disclosure we
provided all we have."
2d UPDATE: Chris now is home and begins the full report. This is part one. It was a 2-hour hearing:
Quite a few lawyers turned out for today's hearing. For IBM, David Marriott, Amy Sorenson, Jeremy Brown, and a couple others were present with David Marriott arguing. For SCO, Stuart Singer argued with Brent Hatch, Ryan Tibbets, and others present.
There was a short discussion of what was to be heard, and David Marriott yielded the podium to SCO's Stuart Singer to present why Marc Rochkind's declaration should be accepted. Stuart said that IBM elected not to submit an expert declaration with their original motion, but instead filed it with their rebuttal. The only way SCO could reply to the declaration is as they have done, that it is perfectly proper.
David Marriot replied that despite what the date of service says, they first received the declaration less than 24 hours ago. SCO said something in its opposition papers that we believe was wrong, he said. Professor Davis only created a table summarizing the versions, files, and lines of code identified in SCO's final disclosure. Rochkind's declaration is on something entirely different. IBM's reply brief does not raise any new issues while Rochkind's does.
Judge Wells asked "How does it hurt you?" to which David Marriott replied "It doesn't hurt us in the slightest way. As long as we have an oportunity to reply it doesn't hurt us." Judge Wells asked why IBM did not submit the expert report with their motion. David Marriott said that it was only when SCO denied what they'd asserted in their motion that they had to rebut it. Otherwise it would just be he-said, they-said. So they needed an expert.
Stuart Singer said that the Rochkind declaration says that in Methods & Concepts the specificity that SCO provided in their final disclosure is sufficient.
Judge Wells said that she will allow the declaration to be submitted and that Dr. Davis can reply within 10 days. (Later the 10 days was clarified as being 10 business days.)
So now we can see why Wells allowed the declaration -- IBM said it doesn't hurt them, so she would naturally take the safer choice, to allow it. In general, you allow the plaintiff every opportunity to make a case. Had IBM argued that they would be harmed by allowing it in, it might have been different.
What we learn from what Chris is reporting is that this is all about methods and concepts, apparently. That is evidently what SCO's expert is testifying about, something that SCO didn't indicate on its list. That is what I suspected when I saw the list, actually, that they were trying to keep from revealing that this is what their case is really about. I suspected it because otherwise there isn't anything there on the list to shake a stick at. By bringing its motion, IBM flushed SCO out of the bushes, and now it's on the table.
And now, all the details from Chris:
David Marriott spoke first on IBM's motion to drop the 198 items on SCO's final disclosure, which IBM believes fail to provide the requisite specificity. He said that he had four main points -- the first that the court specifically ordered SCO to identify "version, file, and line" of code; second, that SCO lacks the required specificity; third, that as a practical matter IBM is at a disadvantage; and fourth, that the only remedy is to not proceed on these 198 items.
David Marriott provided a binder to the court for reference and indicated by reference to tab 1 that the chronology of this motion is long, having started in 2003, only one or two months after SCO filed the lawsuit. He refers to tab 2 and says that SCO contends IBM "dumped" AIX or Unix code into Linux. IBM has been requesting this information for three years. He says that Unix SysV has 11 versions, 112,000 files, and millions of lines of code. AIX has 9 versions, some large number of files, and 1.2 billion lines of code. Dynix has 156 million lines of code, and Linux has over a 100 versions and over 1 billion lines of code. Thus SCO potentially implicates over 2 billion lines of code.
He refers the court to tab 3 and reminds Her Honor that she has repeatedly ordered SCO to identify, with specificity, versions, files, and lines of code. Never in this case has SCO asked for relief from any of these orders. The orders were crystal clear. Mr. Marriottt went over in detail the wording used in the cour's orders to fully, with specificity, and in detail, identify the code. It was so ordered four or five times.
In his second point, he said that SCO's final disclosures lack the required specificity, that it is indisputable that for these 198 items the version, file, and line information is not there. Referring to tab 5 (Addendum 5 to IBM's reply paper, the chart by Professor Davis), Mr. Marriott demonstrates how the chart shows the disputed operating systems AIX, Dynix, Unix, and Linux. He says that a number of these 198 identify Linux versions or files, but they do not provide information on how AIX, Dynix, or Unix are infringed. He said that Rochkind played the primary part in assembling SCO's final disclosure.
His third point was that as a practical matter, IBM is at a disadvantage to fairly defend itself. He compared it to SCO pointing to a giant haystack of 2 billion lines of code and IBM being told to find 198 needles. The original allegation was that IBM dumped lines of Unix code directly into Linux. He went over the types of things IBM would have to determine, such as who wrote the code, is it copyrighted, is it public domain, was it accepted in a patch, it is protectable, is it an unprotected idea, method, etc. He gave as an example an item that SCO did identify with sufficient specificity, that IBM was able to investigate it fully, and on that item was able to demonstrate that it was publicly available and published by Intel, line for line, in 1989.
His fourth point was that the only appropriate remedy is that these 198 be dropped. Judge Wells asked why is this appropriate rather than at summary judgment. David replied that with this it would be very messy at summary judgment, that SCO has not provided enough information to work from. He quoted Judge Wells' statements at an earlier hearing where she admonished counsel about withholding information.
Stuart Singer responded for SCO and provided another book of material to the court for reference. He said that some of this material is confidential but that he believed he could refer to it in this manner. He said he has three arguments, independant reasons, to deny IBM's motion. The first that SCO has complied by disclosing misused material but that there are differences in specificity for methods and procedures. Second, that IBM has no evidence SCO has willfully withheld information, which would be required for discovery sanctions. I didn't catch Stuart's third reason.
On the first, SCO said that they have identified with specificity, but that with technologies the code is not required. Judge Wells reminded him that the orders were to specify files, lines, but that now SCO's talking "technologies". Stuart replied that when SCO was talking about code, they were talking about methods and concepts. Judge Wells says that they have not argued that until now. Mr. Singer said that methods and concepts don't require file, line. Judge Wells said she doesn't remember ever talking about methods and concepts, that we've always been talking about the code. Judge Wells said, "Look at the specific orders". Stuart said they've complied where appropriate, that there has not been a requirement for SCO to identify files and lines for methods and concepts. Where they related to code, they have been identified. He said the contracts between SCO (sic) and IBM required it to keep methods and concepts confidential. He said IBM argues that if it's not source code with file, line, and version, it's not actionable.
He referred the court to tab 7 and provided a summary of the Rothkind declaration. In 130 of the 198, it has the "disclosure" itself. For 168, the file locations were specifically provided.
He drew the court's attention to item 146, indicating it's for a method and concept. It references an article by "McKinney", an old Sequent employee, and points to an IBM web page that they don't have access to. The article is about "Differential Profiling" and refers to scripts on the website. The item has emails between the person making the Linux submission and an individual familiar with the methods and concepts of the protected code. "So we have the identities of the individual making the disclosure, the emails, and reference to code for the methods and concepts."
He said that to throw these out as discovery sanctions, there must be proof of willful withholding of information. There is no such information. What we have is a list of items IBM would *like* to drop. He quoted some cases. He said that if IBM wants to, they can bring summary judgment motions. IBM is seeking discovery sanctions when they have actually sent out an order, after the case was filed, directing their employees to purge their "sandboxes" of AIX and Dynix/ptx code. He said that one developer said he had actual Linux code that he purged. Judge Wells says "That may or may not be true, but I want to stick with things that are properly before the court."
Mr. Singer continued with some more discussion of the items. He said in item 52 (2?) "Method and Concept about Improving Locking," IBM employee "Wright" sent an email to "Daniel Phillips" explaining the "classic Dynix method" and telling Daniel that this is not something currently in Linux. They indicate they have an admission from Wright in his deposition that this came directly out of Dynix/ptx.
From item 38, methods and concepts, SysV "Automatic Method of Making Updates in Memory" he mentioned a specific email with express reference to SysV release 4 for method and concept.
Item 22, "Multiprocessing". Mr. McKinney, again, with a specific email, deposition testimony, and admissions of disclosure of Dynix/ptx methods and concepts.
He pointed out item 23, for "Negative Know How", saying it's a disclosure of how *not* to do something. Where "Martin Bligh", in an email, is telling someone "We are trying to get Linux to have the benefits of (blank) without the pain." That he provided the specifics.
Mr. Singer said that the IBM motion should be denied.
Judge Wells then asked Mr. Marriott to address the "willfulness" issue in his remarks.
Mr. Marriott said that what he heard from SCO was a presentation that had little relation to IBM's motion. He says that the 198 are all deficient.
His first point was that the court had repeatedly ordered lines of code. He agrees that SCO has said methods and concepts, but that the orders applied equally to methods and concepts. He said methods and concepts exist as they are implemented in code. They do not exist in the "ether." SCO must describe, in detail, the location of any method. Location means file and lines of code.
He said that Mr. Singer did not refer to IBM's table (indicating the 198 and files, lines, versions, etc). He said the table is indisputable. The case is about code. We've been pointed to a bunch of our own documents and emails. SCO said, "Here, you know what you did, talk to your own developer, he knows what he did." But that they don't identify the allegation.
Mr. Marriott talked at length about how IBM needs the specifics of the allegations SCO claims IBM has done. That they must have them to know what they are to defend themselves from.
SCO claims that we've disclosed UNIX SysV release 4 internals, Marriott continued. He held up a book ("The Magic Garden Explained, The Internals of Unix System V, Release 4") copyrighted 1994. He said Unix internals have been in public for over a decade. He reiterates that IBM needs specificity; otherwise they are severely prejudiced in preparing for summary judgment.
Stuart rebuts. He said that IBM uses an assumption that for every method and concept, there's source code. He claims that not supported in case law.
Judge Wells asked: "Does SCO have, can they provide, additional specificity?... I mean, basically, is this all you've got?"
To which Stuart replied: "Since, the last two months, it's possible we've discovered more, but, yes, at the time of the disclosure we provided all we have."
Judge Wells asked about SCO only providing links to websites in their final disclosure. Stuart said, yes, it points to a password-protected IBM website.
Stuart said that they have provided the requisite specificity in their disclosures and that there is no evidence of willfulness.
Mr. Marriott then mentioned Judge Kimball's order about dates to disclose by. He said that methods and concepts exist in their implementation in code, not just in the air.
He reminded the court that AIX and Dynix are IBM's product. SCO claims that AIX and Dynix are derivatives of Unix and therefore they control it. SCO has not provided any lines of Unix System V code they say it is derived from.
He said that SCO is the master of their allegations, that they must disclose exactly what is being alleged. IBM cannot do that.
Judge Wells asked Stuart Singer: "How do you address that you maintain custody of the allegations and not provide them to IBM?" Stuart denied that and said that if there is something new that SCO brings up, IBM can object at that time.
Judge Wells concluded the hearing saying that she will take it under advisement and reminds that IBM is to provide expert rebuttal within 10 business days.
That's it. I really couldn't get any feeling from the Judge how this will go.
I think we can see the dance now. IBM brought the motion for a purpose, the stated purpose. If it wins, fine and dandy. If it doesn't, it has won the following: we know now that the case is about methods and concepts now and pretty much nothing else. Think back to what SCOfolk told the world the case was about in 2003 onward, and see if it matches.
We know that at least one item that SCO did mention with specificity on its list was proven already to be bogus.
We know now that the judge, who has seen the Rochkind Declaration, said to SCO, "Is this all you've got?"
And we know now that Rochkind, who is not, as I've pointed out, a Linux expert from what I've seen him say about himself, was the person who helped SCO make up its list, which was a list essentially of Linux files.
And we know that Wells gets it, that this methods and concepts claim isn't what SCO told the court the discovery they claimed they needed was for. Indeed, if what SCO today said is true, that you don't need lines and files and versions if the claim is about methods and concepts, why did they ask for all that AIX code in the first place? IBM was put to great expense, and the case was delayed, and IBM might just like SCO to have to pay for all that, if it was just a fool's errand.
And we've heard SCO's attorney hint that they may introduce more "evidence" which I predict they'll claim to have just discovered in the last two months. I gather they feel their only hope is surprise.
IBM with this motion forced them to put at least one significant card on the table. You are not supposed to have to go to such lengths just to find out what the claim against you is, but Utah is Utah and SCO is SCO, but the bottom line is, SCO was forced to tell what they are suing about, not yet with sufficient specificity, but if the court allows their list to remain intact, SCO will surely have to explain its claims. At last.
The only question now is whether Wells understands the tech enough to understand what David Marriott told her, that methods and concepts are found in the code, not in thin air, which is all SCO wanted to put on the table. Because IBM gets to answer the Rochkind declaration, with Randall Davis, no doubt they'll explain all that very, very clearly, with cases and charts or whatever is needed to make sure the tech is as plain as day to Judge Wells.
And I hope they remember what SCO said to eWeek, not to the court, about Harry Potter, Vanilla Ice and their theory of copyright infringement including methods and concepts. A reminder:
McBride: A lot of code that you'll be seeing coming on in these copyright cases is not going to be line-by-line code. It will be more along the lines of nonliteral copying, which has more to do with infringement. This has more to do with sequence, organization, which is copyright-protectable. It's interesting when you go down this path that everyone wants to go to the exact lines of code, but most copyright cases…
are not line-by-line, exact copies. It's too obvious. Most copyright infringement cases come from these nonliteral implementations of the same code or literary work.
Sontag: My favorite example is the Russian author [Dmitry Yemets], who lost in a copyright case [after being sued by] J.K. Rowling, author of the 'Harry Potter' books, in a Dutch court. He had written a book: It was a girl, not a boy, with magical powers who rides a magical fiddle and not a broom, goes to a boarding school to learn witchcraft and wizardry, plays a game of throwing balls through hoops. All these things were very similar to Harry Potter. Could someone else ever write a book about wizards and witches? Sure. But when the structure and sequence is the same…maybe the words, the code, isn't exactly the same, but Linux is trying to be just like Unix System V. The question is whether Linux was trying to be like Unix System V by doing it in ways that were illegal.
McBride: Before all of this is said and done, you'll see people saying that SCO already published a lot of this stuff in books but that these books contained copyright-protected materials....
No, that's (Lions' book) ancient stuff. We're talking about recent stuff posted as a result of the BSD [Berkeley Software Design Inc.] settlement. There are things out there that help people understand how to program to System V application binary interfaces [ABIs], to help them hook up to the OS. It was out there to help people write applications. It wasn't published to help someone knock off the OS and create a free version of System V. ... The analogy I like to use is Vanilla Ice's "Ice Ice Baby" versus David Bowie and Queen's "Under Pressure." If you just look at the words, I don't see a copyright violation, but if you listen to the riffs, you can hear where they're the same.
SCO told the court that IBM has no proof of willfulness. Please note that the eWeek interview was in May of 2004.
4th Update: I asked Chris if he had any idea what was in the Rochkind Declaration. Could he see it? And yes, he could, and it was very much like IBM's chart, but with different categories. Here's what Chris says:
The Rochkind declaration seems to be a table, not unlike IBM's, but
that has different information on each of the things. It appears to
specify what SCO *did* include for each allegation and whether it meets
the requirements for a Methods and Concepts allegation (with, of course,
lots of filled in check boxes).
The table, on a couple of plain sheets of paper, was held up for
reference by both Stuart Singer and David Mariott. The table's columns
were explained. It looks so similar in physical appearance that David
Marriott at one point mistakenly held it up as if it was his own table,
then corrected himself saying "refer to *our* table that looks *just
like* their table."
But the columns are not Unix, AIX, Dynix, Linux etc. The columns are,
I believe, things like "Admission", "Disclosure Included", "Meets
Req'mts for Methods & Concepts", & etc. (Those are not the titles, but
something like this.) So while it's not really like
IBM's, it was SCO's method of showing that though those allegations
might be lacking in source code, they contained sufficient evidence, in
SCO's opinion, to meet the requirements for specificity and therefore
should not be thrown out.
Marriott pointed out that SCO's allegations lack a "real" allegation.
Alleging that IBM put *their own* code, methods & concepts, whatever,
into Linux doesn't bother them in the least. What SCO was missing was
how this relates to anything SCO owns. SCO didn't say "This method and
concept comes from Unix Sys V, R4, file xyz, lines 67-278 and can be
shown to be implemented in Linux version 2.4.16, file xyz_1.c, lines
120-400 and this admission/submission/email/whatever shows IBM caused
it to be put there." That's what's missing from SCO's final disclosure.