So they didn't finish today, evidently, after all, so there will be more tomorrow. And there are some filings in the SCO bankruptcy as well, which we'll have up shortly. Our eyewitnesses should be filing after that. Is this fun or what?
Here are all the filings in the bankruptcy, and as we retrieve the PDFs, I'll put them here:
In short, everyone wants money from SCO. Tanner wants money but it also wants to be approved to audit SCO's 401K for 2007. Then it'll get mo' money. We'll all have to see if SCO has any, of course, after Novell gets done with them in Utah. There's a hearing on that Tanner motion, I see, on June 17.
This highlights the need for more applications for lawyers, by the way, on GNU/Linux. I know in time it will get there.
Day 3 of the SCO v Novell trial started with SCO's Edward Normand calling William Broderick, SCO's Director of Software Licensing. He said he is responsible for all contract licensing. He was first employed under Unix System Laboratories and continued his employment under Novell, Santa Cruz, and SCO (with the exception of a brief period with a different company).
Mr. Broderick discussed various definitions, software license types (binary and source), and what software versions were licensed under his tenure.
Mr. Normand showed a Unisys Corp. software supplement 202 executed with Unix System Labs in 1991 (pre-Novell). Mr. Broderick read out the fee of $455,000. Mr. Normand then showed a series of software licensing order forms with their associated Prior Products appendixes spanning a wide period of time and executed by Unix System Labs, Novell and Santa Cruz. Mr. Broderick explained they look the same because in transition meetings from Novell to Santa Cruz they were told to only change the name on forms.
Next, he showed a fax to Steve Sabbath at Santa Cruz from Mr. Ackerman of Novell. He highlighted where the fax confirmed "standard software agreement and sublicensing agreement currently used by Novell except with company name and address changes." Mr. Broderick agreed this is in accordance with his understanding.
Mr. Normand displayed Amendment 1 to the APA and highlighted wording that SCO could not enter into new SVRX licenses except in accordance with terms of that Amendment. Mr. Broderick said SCO could not do anything that jeopardized the revenue stream to Novell. He expanded that they could not go to an existing company currently paying $100 and say "execute a new license with us and pay $50".
Mr. Normand displayed NCR Corporation's Supplement 112 License Order Form dated March 1997 with a list of prior products. Mr. Broderick said that the "lion's share" (over half) of the prior supplements would be for additional CPUs or changing designated CPU from one computer to another.
Mr. Normand asked if Mr. Broderick has any way to estimate how many licenses he has been over. He answered several thousands. He said that virtually all (43) had prior products. How many don't show prior products? 8. Alps' prior products list had been intentionally omitted. The other 7 were UnixWare. He explained that the software agreements were getting too long to be "putting in front of customers." He was asked to cut down the number of pages in the license and wanted to put the focus on UnixWare, that they only wanted to see UnixWare on the license. He agreed that if any of those licensees had asked for a list of prior products they would have given them one without a fee.
Mr. Normand showed a license agreement executed by Nihon SCO Japan (couldn't reliably make out licensee's name, but I infer it was Alps). It showed a fee of $375,000 which Mr. Broderick represented was based on number of CPUs licensed. Then he showed exhibit J of Prior Products with text reading it had been intentionally left blank.
Next he presented a letter sent to Unisys in 1996(?) and showed its software Right to Use fees & per CPU fees of $375,000, then its exhibit J list of prior products. He showed the court a graphic (not visible to us) reportedly showing a side-by-side comparison of those two agreements and asked Mr. Broderick what is different. He said one was blank but that they received the same Right to use prior products.
They discussed two more agreements, one with Super Computers International and the other, Lexis Software Corp. It was indicated there had been no extra charge for prior products.
Mr. Broderick agreed, when asked, that SCO gets all UnixWare fees.
It was getting hard to follow testimony in this section as Mr. Broderick was speaking softly.
They discussed audits made by Novell. Mr. Broderick stated that during these years Novell did not ask about UnixWare revenue. When asked what would happen had they asked, Mr. Broderick said that it would have "gone all the way upstairs."
Did anyone try to distinguish between UnixWare licenses and SVRX prior products? No, it was "incidental".
Nest, they discussed why an OEM might license source code. He said that some OEMs developed their own (based on the product) and others simply packaged SCO product.
How was it handled if an OEM used large sections of SVR4 code and only a couple of lines of UnixWare code? Mr. Broderick said they use the "one line of code" rule, that if it contained just one line of UnixWare code, the product would be considered a UnixWare product. He said there was no way they were going to get into discussing using 5% of UnixWare and such. They didn't want to get into that.
They discussed the problem presented by Unisys' Clearpath product. It was described as having 3 separate computers within one box where each ran a different OS. One ran Unisys' own proprietary OS performing certain functions, another running SVR4 MP, and yet another running UnixWare. They explained there was some discussions about how it gets licensed. They proposed that for the UnixWare component that SCO get all the $100,000 list price (with none of that product's Prior Products SVRX going to Novell). Then they'd apportion an amount for the SVR4 MP system to Novell. Novell reportedly agreed to this. A letter was shown from Mr. Broderick to "Cindy" at Novell. The letter quoted the APA language where SCO is required to get Novell's agreement.
They next discussed the 2003 Microsoft Agreement. Mr. Broderick said that Microsoft did not have a pre-existing license. He explained what section 4 of the agreement provided. When asked, Mr. Broderick said that no one else had received an OpenServer source license. He said that it was very important to Microsoft. He said that if you want Unix on Intel you came to SCO. He further said that Novell would not get any revenue from this, as the APA was very specific. He described the 2003 OpenServer installed base as large, that the OpenServer source code was worth more than Unixware since UnixWare source was licensed broadly but OpenServer was closely held.
Mr. Normand showed the 1994 Sun license agreement that stated System V 4.1 was deliberately omitted. Mr. Broderick explained that at the time of this agreement Sun already had a license to this software. SCO produced exhibit 414, not yet admitted, and new to Novell. When shown, it was described as a license between USL and Sun that licensed SVR 4.1 ES. Mr. Broderick explained the reason it was excluded on the 1994 agreement is that it was for the source code used in Sun's product, which was based on UnixWare 4.0.
When Mr. Broderick was prompted to tell of his experience with reviewing existing licenses, he related at length his year-long project to convert an aging license agreement image repository containing 30,000 records to a newer method. He related how he examined every license issued and hand typed this into a 400 page Microsoft Word document. Then he went to Lindon to continue entering information from their records as well as sending the document to the Japan office for additions (who sent back a huge spreadsheet). He said by the end of the year he had 420-some pages of everyone who had licensed Unix and what they'd licensed. This document was then entered as exhibit 367. When asked if there were any material errors in the document, he related that he had recently found some from back in 1993 and corrected them. He said he'd put a lot of work on it and believes it's pretty accurate.
Mr. Normand presented a chart of Unix product releases, by product, by year. Mr. Broderick said it depicted the fall of previous release volume with each new release. He said that people jump on board each new release. The transition point was pointed out of APA SVR4 era to UnixWare era.
An email was presented from Novell's Larry Bouffard to "Washington DC" with the subject "Government Contract Info for SCO". Highlighted was the wording that all info, contracts, assets etc. pertaining to the UnixWare business and the old Unix source code business had been transferred to SCO. "They have bought it lock, stock, and barrel."
Mr. Normand projected an unknown paper document and its attachment 1 (continued). Listed were various software releases spanning from SVR4.1ES/3B2 to SVR4.2MP/UnixWare. With information provided by Mr. Broderick, Mr. Normand notated the page with the Right to Use list prices for each, then totaled to $700K.
Mr. Normand then similarly projected the 2003 Microsoft Agreement Exhibit C titled "Additional Assets". Some products were depicted in italics. Mr. Broderick identified those as releases SCO did not have copies of. Mr. Normand went through the products, notating them with the list fees provided by Mr. Broderick. The total was stated to be $1.249 million with the italicized total as $700K.
Novell's Mr. Melaugh cross-examined him. He first presented NCR's Supplement 112 Licensing Order. Mr. Broderick confirmed that of the preceding supplements, roughly half were changes to designated CPU used or to number of CPUs. He testified that sometime during Novell, the policy was changed to number of designated CPUs rather than specific CPUs. He confirmed Mr. Melaugh's question that it was important to track the source code. He confirmed it was something everyone involved was very careful of. Mr. Melaugh asked what would happen if someone came to him and asked to put source code on any computer in the company. Mr. Broderick stated that they could put it on any CPU, clarified as any one CPU. Mr. Melaugh asked him about putting it on more than one, to which he replied that they would have to execute an additional agreement.
Mr. Melaugh further asked if they would have a right to distribute that source code. No, they would not. He asked if the right to distribute had commercial value. Yes. Was it something they were very careful about? Yes. Mr. Broderick confirmed that none of those prices identified given before granted any ability or power to distribute. Nor do any of those licenses enable the licensee to open source the code. Mr. Broderick confirmed that if a binary was created from that code and distributed, a per-copy royalty must be paid. He agreed that that royalty fee was not included in the numbers Mr. Normand provided.
Mr. Normand displayed a May 26, 1996 letter from Mr. Broderick to Cynthia Lamont at Novell. In it Mr. Broderick stated an APA need for Novell's agreement. He sought Novell's approval to ship SVR4.0MP binary. Mr. Broderick confirmed Mr. Normand's assertion, despite the fact that Unisys already had newer licenses listing SVR4.0MP as prior products. Mr. Normand asked if this was something to do with "new" versus "existing" licensees. Mr. Broderick said he didn't understand. After clarification, he responded that this was because they are separately licensed products. He also said this was an existing licensee at the time of the APA.
The Unisys supplement 233 dated May 28, 1996 signed by Mr. Broderick was displayed as well as its exhibit I, Prior Products. Mr. Normand asked if it's true that at the time Unisys entered into this agreement they had pre-existing licenses. The answer was that he was not sure, but probably. Mr. Normand pulled up Unisys supplement 230, dated October 20, 1995, a list of prior products, exhibit J, prior products carried over into tge next supplement. He showed it side by side with supplement 233 and asked if in 233, Unisys had already paid for prior products. Mr. Broderick answered No, they paid for additional RTU and the prior products came with it. Mr. Normand followed up with another question: at that time, had they paid for prior products? More than likely, was the answer.
He then displayed NCR supplement 112 from March 1997, the prior products list, and asked if, by way of previous supplements, NCR already had all rights or previous products? Mr. Broderick answered that it's likely.
He then showed Stratus Computer's supplement 87 from March 17, 1997(?) and entered into a discussion of how licensing worked. Specifically, Mr. Broderick said that if, per se, you had a UnixWare 2.1 license, with prior products right-to-use, a licensee could access SVR4.1 for code, but that the build made must be a UnixWare 2.1. He stated they would not be authorized to create a derivative work of that prior product, that the license grant was only for UnixWare 2.1. Mr. Normand questioned Mr. Broderick further on this, and he confirmed that even though the licensee had a prior release listed and had a right to use, they would not be permitted to create a derivative of it, since they didn't have a stand-alone grant for that version.
They then discussed several old supplements from 1987-1988 that did not have a prior products list. Mr. Broderick said that these were "just upgrade" licenses, that frequently they would be executed with just a single page.
Mr. Normand conducted redirect asking "I expect each licensee had a supplement 1"? Yes. "And did any of these have a Prior Products list?" Yes. He stated that a licensee could use any code from previous releases, but when code generated, it must be treated as code of the executed license.
Novell's Mr. Melaugh conducted re-cross and asked about the NCR supplement. He asked again, there could be "no derivative product of a prior product?" Mr. Broderick said that was correct.
Mr. Broderick was then finished and excused, and SCO's Mr. Gonzales called Jean Acheson. Ms. Acheson stated she has been controler for SCO since September of 2007. She started with USL in 1990/91 handling accounts payable, then was made Revenue Contract Administrator. She continued with Novell and took over management of binary royalties. For Santa Cruz, she handled worldwide revenue and the same for The SCO Group. She stated that she has tracked and reported Unix revenue the whole time.
When queried, she stated that Novell had no interest in UnixWare revenue. She said that the APA section descibes (SVRX) royalties and how to calculate them. She described her understanding of various terms such as "prior products." She stated that no one was required to pay a fee for prior products.
She was shown a cover sheet of a report of cash collected during January 1997. She described it as stating $4.5/$4.6 million to Novell. She described a few of the report's pages.
Regarding UnixWare revenue, she told the court that Novell never told her to report UnixWare revenue or to report prior products revenue.
Then discussed was an April 26, 1996 letter from Jean to Cindy Lamont, with the subject "Cray source" that memorialized Jean's judgment that Cray's source code sublicensing fees should be retained by SCO. Mr. Gonzalez then showed a subsequent email from Novell to Santa Cruz. Jean Acheson described it, saying that Novell agreed the source code right-to-use fees would be retained by Santa Cruz. It provided a reference to APA amendment 1. She stated that Cray has been a licensee since AT&T days and were paying a few hundred thousand dollars per year.
When questioned about a software license order form, she said that Novell never requested UnixWare licensing revenue and that they never paid any to Novell.
She was asked about her experience with audits. She said that as a public company, they are audited quarterly, and annually they audit customers, by Microsoft, and by Novell twice. Describing Novell's audit of 1998 when they were Santa Cruz, she said they were extremely thorough. Their only request for information on UnixWare was pertaining to the 40% clause. Novell never requested payment for UnixWare licenses. Novell never requested for UnixWare license revenue for licenses granted by Novell.
Next displayed was the 1995 Unisys UnixWare 2.0 license executed by Novell. It showed a quarterly monthly report of Unisys' royalties. Mr. Gonzales used 1Q99 as an example. Ms. Acheson confirmed it indicated $159,966 paid to Santa Cruz. She stated Novell never asked for this revenue in either of their audits.
Ms. Acheson reported that in 1997 the revenue from OpenServer versus UnixWare was about 70/30 and that in 2002 it was about 2/3rds.
Novell's Mr. Melaugh cross examined. He showed the 2003 Sun Agreement and specifically the 2nd page of attachment 1, entitled Description of Technology, Additional Technology. He highlighted System V 4.1 BS/3B2 and asked if it was subject to royalty. Ms. Acheson replied that it is when it is separately licensed. She provided the same answer for several others listed.
He next showed NCR's supplement 112's Prior Products List. Ms. Acheson stated that Santa Cruz was not obligated to pass on royalties. She said that NCR pays royalties on what schedule they are under. If they upgrade from an older one to a newer UnixWare license, then they pay royalties under the new one. This determines the royalties passed on to Novell.
SCO's Mr. Gonzalez then did redirect. Ms. Acheson agreed licensees are buying the latest and greatest release and pay royalties based on that.
SCO's Ted Normand then told the court they next called Jeff Hunsaker, Chief Operating Officer of SCO, formerly Vice President Worldwide Sales and Marketing. SCO's Jason Cyrulnik questioned him. Mr. Hunsaker said that in 2002/2003 SCO was primarily selling OpenServer and UnixWare. He said they accounted for 95%+, 98%+, of revenue. Questioned about OpenServer's installed base, he replied they usually state that they have sold 2 million servers.
Mr. Hunsaker described his usage of OpenServer, Unix, UnixWare, and System V. He described his interaction with customers from which he gains understanding of how commercially valuable the product is. He says he does not recall any customer asking to purchase pre-UnixWare releases.
When asked about SCOsource, he reported no direct involvement. He does recall a conference call discussing where they'd learned of a customer in Kentucky who had unbundled the SCO libraries and were using them in Linux. He was concerned about this. He says they can't do that.
Mr. Cyrulnik showed PowerPoint slides of a SCOsource presentation entitled "SCO System V for Linux" with the text "SCO's shared Unix Libraries from OpenServer and UnixWare for use with Linux." Mr. Hunsaker said they did not use older System V, only the laltest releases. He says that the OpenServer and UnixWare licenses did not allow the libraries to be unbundled.
Mr. Cyrulnik showed a July 31, 2003 email from Mr. Hunsaker to Darl McBride, Chris Sontag, Kevin McBride, and cc'd to many which is entitled "SCO Source Issues and Buyoff." The email reads in part that "only right this license provides is for Linux binary run time copies" and "No connection between UnixWare/OpenServer and the SCO Unix IPC license -- they are independant." Questioned on what he meant, Mr. Hunsaker said he's thinking of things in terms of a binary, a product. Very much different, but they use the same technology.
He next showed a reponse to that email from Kim Jenkins at Morgan Keegan, described as a consultant for the company. The email reads in part "This license does not allow users to run Linux legally. The license allows end users to be clean with SCO."
Next shown was a December 11, 2002 SCO press release entitled "SCO Establishes SCOsource." Mr. Hunsaker discussed why a new division was created.
Discussing an email from John Maciaszek and especially the attached document regarding pricing of SCOsource licenses, Mr. Hunsaker said that they set the price the same as a single-processor copy of UnixWare 7.1.3.
Novell's Mr. Jacobs then cros examined him, first asking about what he does in sales. Mr. Hunsaker agreed with his characterization that he provides "Value for Value" and sells products to enhance the customer's business. Mr. Jacobs asked, "SCOsource was a pretty unique product wasn't it?" Mr. Hunsaker said he didn't understand. Mr. Jacobs clarified that "it didn't enhance their business; it was to protect SCO." Mr. Hunsaker said he didn't know the answer.
Mr. Jacobs asked, "Is System V release 4 valuable in relation to SCOsource?" to which Mr. Hunsaker said he doesn't understand, that he didn't sell SCOsource. He said that customers want the latest technology. There was some discussion as to whether SCOsource was different.
Mr. Jacobs presented a December 11, 2003 email from Mr. Hunsaker to SCO's executive group. It's titled "Notes from OPS Council Strategy Meeting." Under the heading "2004 Objectives -- The 'Goal' is profit" are the bullet points 4) Emerge as industry leader for IP, 5) Gain Linux Justice, and 6) Launch. Another heading reads "SCOsource - Chris/GB" with first heading reading Litigate with the entries IBM, End Users, and Others. The next heading under SCOsource reads License with the entries RTU, Unix, and IPx. Further down is a text section reading in part "What is our business model for SCOsource and SCO? SCOsource - Take 1500 penguins and create a room in Lindon and line them up and place the company brand on each one of them." It goes on to refer to code such as JFS, Malloc, RCU, etc. Further down the page is an entry entitled 4. Consequences, a) License cost is $700-$1400/server, b) Litigate: $$$.
Mr. Jacobs questioned if the "Litigate: $$$" is showing the plan was to scare the customer into paying the license fee to avoid the large cost of litigation. Mr. Hunsaker disagreed.
Jay Peterson next testified. Details to follow, and at
at the end of the day, Mr. Singer said they have one more witness to call that would only take about 15 minutes, and this would be followed by each party's closing arguments. Both expect to finish up quickly tomorrow morning. They half jokingly said it would be about an hour and a half. We'll see.