decoration decoration

When you want to know more...
For layout only
Site Map
About Groklaw
Legal Research
ApplevSamsung p.2
Cast: Lawyers
Comes v. MS
Gordon v MS
IV v. Google
Legal Docs
MS Litigations
News Picks
Novell v. MS
Novell-MS Deal
OOXML Appeals
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v Novell
Sean Daly
Software Patents
Switch to Linux
Unix Books
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

Click here to send an email to the editor of this weblog.

Contact PJ

Click here to email PJ. You won't find me on Facebook Donate Paypal

User Functions



Don't have an account yet? Sign up as a New User

No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.

What's New

No new stories

COMMENTS last 48 hrs
No new comments


hosted by ibiblio

On servers donated to ibiblio by AMD.

Notice this Notice? ~ by Dr Stupid
Thursday, June 17 2004 @ 08:05 AM EDT

We've been trying to make sense of the Novell-SCO spat over who owns the copyrights. Dr Stupid decided that a good way to determine who is right would be to look at the copyright notices back at the time of the Agreement and the later Amendment 2, and see what the conduct of the parties indicated as to what they thought at the time. I think you will find it revealing.

You may wish to have all the relevant documents handy, so here is where you can find all of them:

The ultimate decision as to who owns the copyright will be made by the court, obviously. But the copyright notices Dr Stupid now presents may convince you, as it did me, that Novell has some compelling evidence in hand to make a persuasive case. Rand McNatt also contributed significantly to the research that went into preparing this article.


Notice this Notice?
~by Dr Stupid

Executive Summary

The clearest and least biased indication of the nature of the 1995 agreement between Novell and The Santa Cruz Operation (hereafter "oldSCO") comes from the contemporaneous statements and, more crucially, actions of both parties.

oldSCO's handling of the UnixWare source code in the years following the deal seem to me most consistent with those of a company that had obtained the right to freely derive from and sell products based on the code, but inconsistent with those of a company that had been granted, or believed they owned, the copyrights on that code.

There is also some evidence that indicates to me that the notices of copyright ownership became gradually obfuscated over time, either by accident or design.

A 1996 Clarification of What SCO Purchased from Novell

Here is a 1996 UNIXWARE/OPENSERVER WATCH, which described itself like this:

 "UNIXWARE/OPENSERVER WATCH (UOW) is produced by UnixWare Technology Group (UTG) Inc., a not-for-profit global trade association with one purpose: to provide a forum in which member companies promote, influence and advance the development of UnixWare and allied technologies."

 Notice this apology from the editor for an unclear earlier UOW:

 "AN APOLOGY, AN EXPLANATION AND A COMMITMENT  First, an apology. In UOW #10, there were several incorrect uses of the UNIX trade mark, for which your UOW Editor takes full responsibility. Also, some of the writing might have confused some readers into thinking that SCO had taken the reins of the entire UNIX System business, and not just the UNIX System licensing and UnixWare businesses SCO purchased from Novell.  Again, apologies from your editor." [emphasis added][Note update at end of article]

A History Lesson

In 1995, Novell and oldSCO announced a deal whereby oldSCO would take over the Unixware business from Novell. Novell may have got rid of the UnixWare business, but in 1995 it hardly looked as if Novell was washing its hands of UNIX per se:

"Hewlett Packard, Novell and SCO jointly announced a business relationship to deliver a high-volume UNIX operating system with NetWare and enterprise services. . . .

"Novell will work with HP to produce a high-performance implementation of its NetWare Directory Services (NDS) and File/Print Services for HP-UX, and integrate NDS with DCE. . . .

"SCO will utilize Novell's Provo, Syndey and Dusseldorf UnixWare support staff through Q2'96.   SCO will fully support UnixWare products with its own staff beginning in Q2'96."

Novell and oldSCO had more than one link:

"Novell will receive approximately 6.1 million shares of SCO common stock, resulting in an ownership position of approximately 17%  (post transaction) of the outstanding SCO capital stock. . . .

"In order to meet customer support needs and protect development requirements, SCO intends to hire a number of Novell employees. "

At the time of the deal, Novell was already working on a new version of the product, which would be officially called UnixWare 2.1.

Although the deal was signed in 1995, the actual handover of the business from Novell to oldSCO would not happen until early 1996. In the meantime Novell kept working on Unixware, and in December of 1995, UnixWare 2.1 was completed.

"[UW2.1 was] 100% [Novell].  The ctime dates on the SCO UnixWare 2.1 CD-ROM (the commercial version) are Dec 12 and Dec 13 1995.  The business was transitioned to SCO on 31 Jan 1996."

In early 1996, Novell handed over the UnixWare business to SCO, including the new product. Unixware 2.1 was released to the public as a SCO product, though no one at oldSCO had actually written any of it.

While oldSCO was getting familiar with their new product, they hired a number of Novell employees to maintain it in the meantime - remember:

"In order to meet customer support needs and protect development requirements, SCO intends to hire a number of Novell employees. "

Late in 1996, SCO released an updated version of UnixWare - UnixWare 2.1.1. This was over a year after the original deal was announced, and 9 months after the handover of the business.

One would suppose, from the above history, that Novell's contributions to UnixWare had stopped in December of 1995. But this is not so; and we can see that because we can see one of the files that make up UnixWare 2.1.1.

A Novell employee, as part of an internet discussion in 1997, revealed the copyright header on one of the Unixware 2.1.1 files - an important "header file" called "stdio.h"

/*Copyright (c) 1990, 1991, 1992, 1993, 1994, 1995, 1996 Novell, Inc. All Rights Reserved.*/
/*Copyright (c) 1988, 1990 Novell, Inc. All Rights Reserved.*/
/*  All Rights Reserved  */

/*The copyright notice above does not evidence any   */
/*actual or intended publication of such source code.*/

#ifndef _STDIO_H
#define _STDIO_H
#ident"@(#)/usr/include/ 1.1 u211 09/27/96 52255 SCO"

The "u211" shows this file is from Unixware 2.1.1. The 9/27/96 is the date the file was last changed - months after Novell had handed over to oldSCO. And the copyright notice includes a (c)1996 Novell - no SCO at all.

So, even months after the handover from Novell to SCO of the UnixWare business, Novell were still making improvements to UnixWare and oldSCO was acknowledging them as holding the copyrights. The Novell employees helping oldSCO with support in early 1996 could not have been "working for hire" for oldSCO as otherwise their contributions would have been (c) SCO, irrespective of the APA amendments. Novell and oldSCO were still acting as if Novell owned the copyrights to the SysV core materials, and oldSCO was the "publisher" to the wider world.

These files were freely available from SCO's ftp server at the time, via anonymous download, and they can still be found at various mirrors for SCO customers.

The 2.1.1 update files confirm the above USENET posting and also contain other files with a (c) 1996 Novell copyright, e.g.:

/*    Copyright (c) 1990, 1991, 1992, 1993, 1994, 1995, 1996 Novell, Inc. All Rights Reserved.    */
/*    Copyright (c) 1984, 1985, 1986, 1987, 1988, 1989, 1990 Novell, Inc. All Rights Reserved.    */
/*      All Rights Reserved      */

/*    The copyright notice above does not evidence any       */
/*    actual or intended publication of such source code.    */

#ifndef _SVC_ERRNO_H    /* wrapper symbol for kernel use */
#define _SVC_ERRNO_H    /* subject to change without notice */

#ident    "@(#)/usr/include/sys/ 1.1 u211 09/27/96 44690 SCO"
#ident    "$Header: $"

By the time of the next update, UnixWare 2.1.2, oldSCO had (according to the plan outlined above) taken over development of the product. Yet the 2.1.2 update (from 1997) still includes non-NetWare material with (c) Novell on it.

The 2.1.3 update is from 1998 and is much larger. It includes many updates to man pages (still with Novell copyrights.) The file
/usr/sbin/pppconf is updated to vintage 1998 - still (c) Novell; and /usr/include/sys/xti.h is also still (c) Novell.

In fact, there are only 2 source files in this update which have a (c) Santa Cruz copyright - a small shell script and this:

 *    Copyright (C) 1997 The Santa Cruz Operation, Inc.
 *        All Rights Reserved.
 *    The information in this file is provided for the exclusive use of
 *    the licensees of The Santa Cruz Operation, Inc.
 *    The information in this file is provided "AS IS" without warranty.
 *    Copyright (C) 1995-1997 Intel Corporation.
 *        All Rights Reserved.
 *    This update binary code is distributed for the sole purpose of
 *    being loading into Intel P6 Family microprocessors in systems
 *    upon which your operating system is installed or executed.
 *    You may not make any derivative work of, nor perform any reverse
 *    engineering upon, the update binary code, nor facilitate the
 *    update to be loaded into any non-Intel processor.
 *    Tunable INTEL_CPUREV is autotuned by idbuild, to minimize the
 *    size of the Pentium Pro & Pentium II microcode updates table.

#include "config.h"

#ifndef INTEL_CPUREV                /* tunable is missing so */
#define INTEL_CPUREV    0x6000000        /* link all p6_updates */
#define CPU        (INTEL_CPUREV >> 16)    /* family, model, step */
#define REV        (INTEL_CPUREV & 0xffff)    /* microcode revision */

unsigned long p6_updates[] =
#if     (REV 0x00000001,    0x00000b27,    0x12181996,    0x00000611,
0x05793e46,    0x00000001,    0x00000000,    0x00000000,
0x00000000,    0x00000000,    0x00000000,    0x00000000,
[snip lots of hex data]
/* record built INTEL_CPUREV in terminating line of table */
0x00000000,    INTEL_CPUREV,    0x00000000,    0x00000000

I'm not sure what that (c) SCO is for, since they seem to have simply taken the Intel file and just put their own header on it.

Here is what is supposed to be in a copyright notice, according to the US Copyright Office's Copyright Basics page:

"Form of Notice for Visually Perceptible Copies

"The notice for visually perceptible copies should contain all the following three elements:

1. The symbol © (the letter C in a circle), or the word "Copyright," or the abbreviation "Copr."; and

2. The year of first publication of the work....

3. The name of the *owner of copyright* in the work, or an abbreviation by which the name can be recognized, or a generally known alternative designation of the owner." [emphasis added]

So, the names mentioned should be the copyright owners, not a history of past authors. Novell wasn't writing UNIX in 1984, AT&T was. But because Novell bought the copyright from AT&T/USL, they got to say "(c) 1984 Novell."

It follows logically that if oldSCO bought the copyrights from Novell, the notice should read "(c) 1984-1997 The Santa Cruz Operation" and not mention Novell at all.

What does this mean?

  • Several months after the deal that newSCO now claims gave them the copyrights, oldSCO was happily selling a product that showed Novell as owning copyrights.
  • Novell was still contributing code to the product even after oldSCO supposedly "took it over".
  • In 1998, years after the APA and years after Amendment 2, oldSCO was still leaving (c) Novell notices untouched in its updates to the UnixWare system - including updates to the manuals of which newSCO explicitly claimed the copyright in its complaint against AutoZone.
You ain't seen nothing yet

As mentioned, the 2.1.3 update does not add SCO copyrights to system files: in fact it contains several updated man pages and some headers that still say (c) Novell. However, it does alter some system header files.

If one compares the errno.h file from UW2.1.1 and UW2.1.3 with a "diff" program, one can see that none of the code has changed. The only real change is that the copyright notice has been removed. That's right - (c) Novell is not replaced with (c) SCO, but instead the copyright header is simply stripped.

The same is done to various other system header files in the 2.1.3 update. The reader is invited to draw his or her own conclusions.

The step from 2 to 7

In 1998 oldSCO released UnixWare 7  - the merger of OpenServer 5 and UnixWare 2 (The "merged product" referred to in the APA.) The following makes use of information obtained from a legitimate UnixWare 7.1.1 system and its associated media kit.

Looking at the actual files on a UnixWare 7 system, we find that the errno.h file contains a simple (c) 1998 The Santa Cruz Operation notice. Of course this copyright notice is not complete since errno.h was not wholly authored in 1998, but the more significant point is the overall history of this file while it was in oldSCO's hands:

  • UnixWare 2.1.1 - (c) Novell
  • UnixWare 2.1.2 - (c) Novell
  • UnixWare 2.1.3 - No copyright notice
  • UnixWare 7 - (c) oldSCO

CD1 - the installation CD - boots to a miniature UnixWare system to conduct the install. The data for the installation is held in several folders on the CDROM, each of which is like a partly unpacked "pkg" format archive. I say partially, because although the individual files are present in a series of directory structures topped with "/root.#" (where # is a number) rather than held in a cpio-style archive, most are compressed and must be unpacked with the traditional UNIX "uncompress" before reading.

From the Horse's Mouth

There is a package called "BASE", which as you might guess from the name contains (inter alia) the UW kernel:

PSTAMP=UW7 04/07/98
NAME=Base System
DESC=Base Operating System, commands and utilities.
CLASSES=sysdir kernel sysutil inst vtool intrfc config term modem sysinst need upnover none
ORDER=sysdir kernel sysutil inst vtool intrfc config term modem sysinst need upnover none

[Aside: the installation program actually builds the kernel (from a collection binary modules) during the installation.]

Here is the copyright info for this package, taken from the "copyright" file which it contains:

(C) Copyright 1996-1998 The Santa Cruz Operation, Inc.  All rights reserved.
Copyright 1984-1995 Novell, Inc.  All Rights Reserved.
Portions Copyright (c) 1989 INTERACTIVE Systems Corporation.
Portions Copyright (c) 1990, 1991, 1992 INTEL Corporation.
Portions Copyright (c) 1993 Compaq Computer Corp.
All Rights Reserved.

In fact there are 96 "copyright" files mentioning Novell on CD1 alone, most of the format:

(C) Copyright 1996-xxx The Santa Cruz Operation, Inc.  All rights reserved.
Copyright 1984-1995 Novell, Inc.  All Rights Reserved.

These "copyright" files are not mere leftovers inherited from Novell, since at the time of the APA they would have contained only the Novell copyright notice. Therefore oldSCO had (in 1996-1997) gone through each of these files and added their copyright notice dating from 1996, quite deliberately leaving the Novell notice intact.

A Cast of Thousands

Now, the install program places a "copyrights.list" file on the hard disk, which reads (in part):

Copyright (c) 1976-1998 The Santa Cruz Operation, Inc. All Rights Reserved.
Copyright (c) 1976-1992 AT & T
Copyright (c) 1987-1995 Computer Associates International, Inc.
Copyright (c) 1988-1997 Edison Design Group, Inc.
Copyright (c) 1983-1997 Eric P. Allman
Copyright (c) 1993-1995 Hewlett-Packard Company
Copyright (c) 1995      Hitachi, Ltd.
Copyright (c) 1990-1998 Intel Corporation
Copyright (c) 1992-1998 International Business Machines Corporation
Copyright (c) 1984-1998 Massachusetts Institute of Technology
Copyright (c) 1987-1988 Microsoft Corporation
Copyright (c) 1993-1998 Novell, Inc.
Copyright (c) 1982-1995 The Regents of the University of California
Copyright (c) 1993-1995 Sun Microsystems, Inc.
Copyright (c) 1989-1997 The Open Group (formerly OSF)
Copyright (c) 1992-1998 Compaq Computer Corporation
Copyright (c) 1994-1998 Digital Equipment Corporation
All Rights Reserved.

A certain inconsistency is apparent with the "copyright" data on the packages themselves; but note the following:
Copyright (c) 1976-1998 The Santa Cruz Operation
Copyright (c) 1976-1992 AT & T
Copyright (c) 1993-1998 Novell, Inc.

The AT&T and Novell ranges together span the years of oldSCO's declaration. Aside: how did AT&T still hold a copyright on part of UnixWare?

Although oldSCO appears to be claiming a copyright going back to 1976, there's only one package with a COPYRIGHT file that mentions 1976. Strangely, it's for the CDE login manager, which I believe didn't exist back in 1976.

The notice starts (emphasis mine):

UnixWare 7

Copyright (c) 1976-1998 The Santa Cruz Operation, Inc. and its suppliers.
All Rights Reserved.

The "and its suppliers" effectively makes the copyright notice very vague about who owns what. It effectively forces one to look at all the other notices - which as we have seen, include several references to Novell. Overall, oldSCO's stance looks like an assertion of collective copyright - that oldSCO had the copyright on the specific compilation and integration of programs and code that made up UnixWare, and could thus pursue those who copied UnixWare as a whole, but had little copyright interest in any given program or code. [The situation is similar to that of the publisher of an anthology of poetry.]

As mentioned previously, the installation program actually builds the kernel (from a collection of binary modules) during the installation. The UnixWare kernel is also rebuilt from those modules as part of certain system configuration procedures. The "svc" module has within it:

UnixWare %v for the Intel386(tm) Family
Copyright 1984-1995 Novell, Inc.  All Rights Reserved
U.S. Pat. No. 5,349,642

None of the other modules have a Santa Cruz copyright that I've spotted yet.

What is the "svc" kernel module for? According to this USENET post, John Wiegley tells us that svc is the "System V configuration" module.

That boot-up message ("UnixWare %v for the Intel386(tm) Family") is the built-in default - it can be overridden by a file called "bootmsgs". Sure enough, oldSCO provided one - here is the relevant extract:

BOOTMSG1=Starting UnixWare...
TITLE=UnixWare 7, based on UNIX System V Release 5 from SCO
COPYRIGHT=Copyright (c) 1976-1998 The Santa Cruz Operation, Inc. and its suppliers. All Rights Reserved.

And Still Footprints in the Code

Along with the operating system binaries, the installation also includes system header files. Here is a particularly interesting one:
/usr/include/machlock.h - it appears to deal with atomic locks, a low-level system operation.

 * Copyright (c) 1999 The Santa Cruz Operation, Inc.. All Rights Reserved.
 *                   SANTA CRUZ OPERATION INC.
 *   The copyright notice above does not evidence any actual or intended
 *   publication of such source code.

/*      Copyright (c) 1990, 1991, 1992, 1993, 1994, 1995 Novell, Inc. All Rights Reserved.*/
/*      Copyright (c) 1988, 1990 Novell, Inc. All Rights Reserved.      */
/*        All Rights Reserved   */

/*      The copyright notice above does not evidence any        */
/*      actual or intended publication of such source code.     */

Let's compare that to another UnixWare header file that  Novell definitely retained the copyrights to (by the admission of newSCO's own lawyers) - the NetWare stuff:

 * Copyright (c) 1998 The Santa Cruz Operation, Inc.. All Rights Reserved.
 *                   SANTA CRUZ OPERATION INC.
 *   The copyright notice above does not evidence any actual or intended
 *   publication of such source code.

/* $Novell-NWU: $Header: /proj6/ncps/nwu_top/nuc/include/nw/nwerror.h,v 1.3 1996
/*      Copyright (c) 1990, 1991, 1992, 1993, 1994, 1995 Novell, Inc. All Rights
/*      Copyright (c) 1993 Novell, Inc. All Rights Reserved.    */
/*        All Rights Reserved   */

/*      The copyright notice above does not evidence any        */
/*      actual or intended publication of such source code.     */

Not enormously different is it? Would you conclude, looking at these two files, that Novell owned the copyright on the latter and The Santa Cruz Operation on the former? Or would you conclude that Novell owned them both and oldSCO was merely asserting its collective copyright on UnixWare as a whole?

A Scrivener's Error?

newSCO has dismissed the original APA's explicit exclusion of copyrights as a "scrivener's error" - that is to say, that it was everyone's intention in 1995 that oldSCO receive the copyrights and the agreement was merely incorrectly drafted. However, on its face the UnixWare 2.1.1 update shows me that in 1996
(a) Novell was more involved in UnixWare at the time than newSCO has subsequently represented;
(b) oldSCO was not acting like a company that had bought the copyrights; 
(c) Novell was not acting like a company that had sold them.

This is not the only evidence, moreover. In particular, if the exclusion of copyrights was really just a typo, then one would have expected it to be corrected in Amendment 1. Instead,
Amendment 1 actually adds the language

 "In addition, Buyer shall not, and shall have no right to, enter into new SVRX Licenses" [emph mine]

There is a second, more subtle variant on the "error" scenario that might be advanced - namely, an analogue of the oldSCO/Caldera deal in which oldSCO originally retained the OpenServer rights and later sold them to Caldera. In this scenario the copyrights were not intended to be transferred, but later on oldSCO and Novell agreed to transfer them for some reason and created Amendment 2 to that effect.

This scenario, though, again fails to gel with the actions of the companies at the time. Surely such a deal would have merited a press release, but there was no such release that I could find. The UW2.1.2 and 2.1.3 updates do not place SCO copyright notices on the core UNIX code; and UW7 has the tell-tale copyright notices which still list Novell as the copyright holder of contributions prior to 1996.

As an aside, Amendment 1 also says:

"1.2 (d) Asset Transfer and Transfer Taxes. Notwithstanding any other provision of this Agreement, the Assets shall remain the property of Seller until expeditiously delivered to Buyer in the manner and at the locations prescribed as follows in this Section 1.2(d), or as subsequently agreed in writing."

and then makes provision for delivery of physical source code but no specific arrangments for transfer of copyrights. This does strongly indicate that any copyright transfers were to be enacted by separate writings from the APA itself.

What, then, was the purpose and motivation behind the amendment?

The trademark patch

Firstly, the original APA gave as excluded assets:

"All copyrights and trademarks, except for the trademarks UNIX and UnixWare."

This apparently gave oldSCO the UNIX trademark. Except this was obviously wrong, since the UNIX trademark had been given to the Open Group. So this clause had to change. Also, oldSCO bought a business that sold many products - the UnixWare trademark alone might not cover them all. Clearly, Novell had to sell oldSCO the trademarks it needed to conduct the business.

That would appear to be the prime mover behind this part of the Amendment. As we see elsewhere, there is nothing in oldSCO's behaviour around this time to indicate that it thought it
owned the UNIX copyrights (and thus would want the APA "brought into line")

However, given that oldSCO was an existing UNIX licensee of an older version (not SVR4), it wasn't clear whether they could rely on that license alone to carry on the business of selling UnixWare. There might be the odd bits of UnixWare that couldn't be counted under the SysV license. All the above combine to yield the final catchall language:

"All copyrights and trademarks, except for the copyrights and trademarks owned by Novell as of the date of the Agreement required for SCO to exercise its rights with respect to the acquisition of UNIX and UnixWare technologies."

What are those rights? To refer blindly to the start of the APA is inadequate since it introduces a hopeless circularity:
SCO rights = All rights - Patents - (Copyrights - (Copyrights needed for SCO rights, which are.... oops))

APA Amendment 1 gives a hint, though, when it says
" may be incidentally involved through its rights to sell and license UnixWare software or the Merged Product.."

How did oldSCO describe the transaction?
"In fiscal year 1996, SCO acquired the UnixWare(R) and UNIX System V Release 4 source-license business from Novell, Inc."

[Another point. oldSCO were to be given those copyrights required to exercise their rights. UnixWare contained NetWare code, yet oldSCO did not require ownership of the NetWare copyrights to develop and sell UnixWare. Why, therefore, should they require ownership of the ancient UNIX and System V copyrights?]

The Real Point of the Amendment?

Secondly, there is more to Amendment 2 than this oft-discussed clause. The bulk of the amendment creates arrangements for royalty buyouts of System V: arrangements whereby both oldSCO's and Novell's permission is required. Why might this have come to the fore? Presumably because in the two previous years there had been two royalty buyouts associated with System V:

"The increase from fiscal 1995 to fiscal 1996 was attributable to a one-time $19 million paid-up royalty recognized in the sale of UNIX technology to SCO in fiscal 1996. The decrease in fiscal 1995 compared to fiscal 1994 was due to a one-time fully paid license for UNIX technology sold to Sun Microsystems for $81 million in fiscal 1994."

It is worth pointing out that this $19m royalty was not mentioned as part of the original deal between Novell and oldSCO. It also strikes me as unlikely that any payment which bought a copyright outright would ever be described as a royalty payment. A "one-time $19 million paid-up royalty" sounds more like a "right[] to sell and license".

Amendment 2 states that
"Novell may execute a [System V royalty] buy-out ... without any approval or involvement of SCO, ... if: (i) SCO ceases to actively and aggressively market SCO's UNIX platforms.."

That seems to say that Novell could seek to generate revenue from royalty buy-outs if oldSCO lost interest in UNIX. This reads not as a right granted to Novell, but as a right held but restrained as long as oldSCO "actively and aggressively" marketed UNIX.

So, the infamous clause in Amendment 2 was not its raison d'etre, it seems. Rather, Amendment 2 could have been drafted primarily to formalize arrangements for subsequent royalty buyouts, like IBM's "Amendment X", and the opportunity was taken to fudge some incorrect wording at the same time.


Based on the above publicly available evidence, which -- being contemporaneous -- is the least likely to be coloured by today's agendas of the parties involved, in my opinion, the deal between oldSCO and Novell was most likely as follows:

  • Novell sold the physical assets, reseller channel and information related to the UnixWare business to oldSCO and was given oldSCO stock in return. This was part of an overall strategy whereby Novell, HP and oldSCO would co-operate on driving forward the Unix platform.
  • oldSCO were given every right they needed to conduct the business -- the trademarks, and the right to develop and license the UnixWare code as they saw fit.
  • Novell retained ownership of  the copyrights and patents, both so as to retain a revenue stream from royalties and for its internal needs.
  • Because Novell retained such broad rights to UNIX, oldSCO needed assurances that Novell would not undermine their business. Thus, Novell was contractually bound not to create new System V licensees, not to try to "sell" System V in general, and so forth.
  • In return, Novell needed assurances that oldSCO would not undermine the System V royalty stream by allowing licensees to buyout their System V licenses, or by harassing the licensees. Thus, Novell retained far-reaching powers of waiver with respect to the old licenses, and it was agreed that royalty buyouts needed both parties' approval.
  • oldSCO negotiated a royalty buyout in respect of its existing System V licenses, so that it could create the merged product (UnixWare 7) and sell it without paying any royalties to Novell.


An enterprising Groklaw reader, Sean Lynch, used his research skills to contact Michael Dortch, mentioned in the article, who has some remembrances to add to the picture. He has given us permission to publish the email, which confirms that the article is "essentially accurate and correct" and that to the best of his knowledge, "SCO only purchased from Novell UNIX System licensing rights and the UnixWare business":

Thanks for your e-mail. I must say at the outset that I was initially surprised to see a quote from "UnixWare/OpenServer Watch" in a current e-mail. It seems that some artifacts never die online.

In any event, yes, I am the Michael Dortch who was the editor of that newsletter at the time the piece containing the quote referred to in the Groklaw article was published. I created that newsletter while at UnixWare Technology Group (UTG), a trade association of UnixWare (and, later, as I recall, OpenServer) resellers and other supporters. I was "chief evangelist" for UTG, recruited by its founding president and CEO, Lawrence Lytel, in 1994. (I am now a principal business analyst at Robert Frances Group (RFG), providers of advice and counsel to IT executives and their teams and colleagues. In this role, my colleagues and I have also been following closely and striving to untangle the details of the "Novell-SCO spat," as the Groklaw article so charmingly describes it.)

Yes, UTG and I were contacted by Grant Bird, then Director of Branding at X/Open, predecessor to today's Open Group. As it says in the issue of "UnixWare/OpenServer Watch" from which the Groklaw article extracted my quote, X/Open produced and "owned" the then-"official" guide to proper use of terms such as "UNIX System," and I had misused some of those terms in a previous issue of the newsletter. It was my discussions via e-mail and telephone with Grant that led to the apology and commitment from which the Groklaw article quote was extracted.

As I recall, my discussions with Grant focused solely on proper use of terms, not ownership of the intellectual property described by those terms. However, I do have some thoughts and recollections on that subject as well.

UTG was formed to help Novell promote UnixWare, and was initially populated largely by Novell UnixWare partners such as Unisys. As I recall, Novell was the primary initial financial supporter of UTG, apparently intending to encourage collaborative marketing among UnixWare partners as well as interoperability among their various respective UnixWare and UNIX System offerings.

HP was also prominently involved; representatives from HP and what you refer to as "oldSCO" made a joint presentation to a UTG member meeting held on March 27, 1996 in Japan, for example. At that event, oldSCO said it had hired more than 200 former employees of Novell and "USL," which was the old AT&T UNIX System Laboratories from which, as I recall, Novell bought the rights to UNIX System V. In any event, oldSCO said the "Next Generation OS" being developed would be based on "SCO's and HP's 3 products: UnixWare, OpenServer, and HP-UX." (FYI, I parted company with UTG in July 1996, and the organization itself was dissolved shortly thereafter. As I recall, there was fairly substantial doubt that HP was really more interested in collaborating on a "Next Generation OS" than in promoting HP-UX as that solution. There was also doubt among UTG's former members that SCO could really lead a charge toward such an OS without full support from HP and/or Novell.)

In any event, based on my recollection of events at the times mentioned in the articles to which you referred in your earlier e-mail, I believe the Groklaw article's interpretation of both my quote and the events at the time is essentially accurate and correct. To the best of my knowledge, as it says in my quote, SCO only purchased from Novell UNIX System licensing rights and the UnixWare business. Also, Novell did in fact continue to work on and champion UnixWare for some time after that purchase SCO was made final.

The vast majority of RFG clients are IT executives at large enterprises. Many of these use or are seriously considering using Linux and other technologies potentially affected by the labyrinthine legal machinations of "newSCO." As RFG said in a March 2004 analysis, "Is SCO At the Tipping Point?," "SCO's case continues to disintegrate, and proving its claims against IBM will be extremely difficult. Until and unless SCO is able to do so, RFG does not see value in paying SCO protection money just to use Linux. ...IT executives should not fall prey to SCO's scare tactics, and should instead continue to focus on the ultimate values and risks involved in each platform decision made as part of an application deployment."

I hope this is helpful, and appreciate your contacting me. Please feel free to quote from the above text in whole or in part, as long as any such quotes are made verbatim and include appropriate attribution. Also, feel free to contact me again with any other questions or comments you may have on this still-fascinating topic.


Michael Dortch
Principal Business Analyst
IT Infrastructure Management Practice Leader
Editorial Director
Robert Frances Group
"Business Advisors to IT Executives"


Notice this Notice? ~ by Dr Stupid | 525 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Corrections here, please.
Authored by: darksepulcher on Thursday, June 17 2004 @ 08:17 AM EDT
Makes life easier on PJ.

Had I but time--As this fell Sergeant, Death
Is strict in his arrest--O, I could tell you--
But let it be.
(Hamlet, Act V Scene 2)

[ Reply to This | # ]

trolls please post here
Authored by: paul_cooke on Thursday, June 17 2004 @ 08:19 AM EDT
so the rest of us can easily skip over you...

Use Linux - Computer power for the people: Down with cybercrud...

[ Reply to This | # ]

OT posts and new URLs here please
Authored by: Acrow Nimh on Thursday, June 17 2004 @ 08:20 AM EDT
Very nice analysis BTW...

Supporting Open Sauce since 1947 ;¬)

[ Reply to This | # ]

The APA and copyrights
Authored by: codswallop on Thursday, June 17 2004 @ 08:26 AM EDT
Given our choices of user names, I feel particularly qualified to post this.

This is a summary of the current state of my analysis of the APA-copyrights
issue and related matters. I'd like to thank the people here and on Yahoo who've
helped with this. Any mistakes are, of course, my own.

The APA as amended says that Novell

"will sell, convey, transfer, assign and deliver to...[The Santa Cruz
Operation] ...on the Closing Date .. all of Seller's right, title and interest
in and to..... All rights and ownership of UNIX and UnixWare and Auxiliary
Products [and the] Trademarks UNIX and UnixWare as and to the extent held by
Seller.... Notwithstanding the foregoing, the Assets to be so purchased shall
not include All copyrights and trademarks, except for the copyrights and
trademarks owned by Novell as of the date of the Agreement required for SCO to
exercise its rights with respect to the acquisition of UNIX and UnixWare

I made this by gluing together the language from section 1.1(a) of the APA with
that from schedules 1.1(a) and 1.1(b). This is a rather ugly construct, but it
is clear enough to suggest a few things.

1) The word required as used here doesn't imply that anything is supposed to
happen after the closing (or, since the amendment was signed after the closing,
the signing of the amendment). There's no support for a future demonstration of
need that would trigger a transfer of rights.

2) There is a set of "required" trademarks and a set of "required
copyrights" that are to be transferred at closing.

3) These are not defined in any useful way. The implication is that both sides
know what these sets contain.

4) Given this language from 1.7(b):

"(iii) Seller shall deliver to Buyer all bills of sale, endorsements,
assignments, consents to assignments to the extent obtained and other
instruments and documents as Buyer may reasonably request to sell, convey,
assign, transfer and deliver to Buyer Seller's title to all the Assets;"

it seems that there was no need to define what the required sets were with
enough specificity to effect a transfer. There were supposed to be separate
conveyance instruments. This also argues against the APA as a section 204(a)
writing. Note that the APA may be a valid instrument for the trademarks, because
the requirements are different.

5) What would be expected would be a schedule of instruments to be brought to
the signing. Given how close the signing was to the drafting of the amendment, a
reasonable conjecture is that it wasn't felt to be necessary, as the issue would
have no significance after the signing and the transfer of the conveyance

6) There are no grounds for saying Santa Cruz couldn't have gotten some
particular set of rights because they weren't "required" according to
some interpretation of "exercise" or "rights". If the clause
doesn't effect a transfer, neither does it restrict one. It's just a descriptive
phrase for the documents Novell was to deliver.

This is what the amended APA says. The harder question is what the parties
thought it meant when they drafted it and when they signed it. There are several
possible theories:

1) Both sides thought it would convey copyrights. In this case the clause is
void, both because it isn't a valid section 204(a) writing and also because the
interpretation is a mutual mistake.

2) One side thought it would convey copyrights and the other didn't. The
section 204(a) problem is still there, but now the mistake is unilateral.
Unilateral mistakes may void a contract section depending on circumstances and
state law. Also in this case if Novell was the nonbelieving party, it was bad
faith for them not to bring the assignments to the signing or at least raise the
issue. Of course we don't know that they didn't do one or both of these things.
In any case this is a cause of action only for Santa Cruz - Tarantella. Since
Caldera-SCO bought only those rights that Santa Cruz had at the closing of their
agreement, and Tarantella survives as a company, it's hard to see what standing
SCO might have in the matter.

3) Neither side believed it would convey copyrights, but for some reason they
didn't care, or decided to ignore or postpone the issue. Complicated, but again
it's unlikely SCO have any standing in the matter.

4) Neither side believed it would convey copyrights, and Novell gave Santa Cruz
what they had agreed to. This case includes the possibility that the set of
copyrights was empty. As was mentioned in the lead article, the trademark
language needed touching up. The addition of copyrights to trademarks could have
been a bit of lawyerly zeal that was later decided by both parties to have no
useful effect.

There are several other facts that have a bearing on this issue:

1) SCO is not a party to the APA either as a signer or a successor in interest.
The APA says in section 9.5:
"This Agreement ... shall not be assigned by operation of law or otherwise
except as otherwise specifically provided"

This means not by ordinary assignment or by merger, acquisition, bankruptcy,
etc. SCO can't sue Novell for breach of contract or for specific performance.
SCO are only the current licensees for the licenses created and/or conveyed in
the APA and probably also the successor in interest to Amendment X, since Santa
Cruz was a signatory. Their rights under the earlier ATT agreements are only
those which they got form Novell through Santa Cruz. Those transfers could use
some closer examination.

2) SCO announced that they were unable to find the copyright assignment
attachments to the Santa Cruz - Caldera agreement. Without these and the
associated assignment documents, they can't demonstrate what copyrights, if any,
they got from Novell by way of Santa Cruz. It's quite clear that the agreement
itself is not a valid section 204(a) writing.

3) The facts that neither the amendment nor the orignal contract convey
copyrights by themselves, that the putative assignment uses the future tense and
that the amendment was signed after the contract don't invalidate the amended
APA as a 204(a) writing. The amendment signed by the parties serves as a
confirming writing at the time of signing, and thus of the conjectured transfer.
There are good arguments for the amended APA not to be a section 204(a) writing,
but this isn't one of them.

[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: kibbey on Thursday, June 17 2004 @ 08:40 AM EDT
Very interesting analysis. Sounds more and more like the good ship SCO is
taking on water in the engine room at an alarming rate.

[ Reply to This | # ]

SEC Filings...
Authored by: kberrien on Thursday, June 17 2004 @ 08:46 AM EDT
Heh, great analysis! Get this gentleman an honorary journalism degree, he's
don't better than most of the pro's.

The other day, on another thread (within the last 5) someone was posting SEC
filings from the tranfers time - which had some interesting admissions from
oldSCO about "buying the business" etc.., and some from Novell about
what they sold.

I don't have time to dig for them now, but might make a good follow up

[ Reply to This | # ]

Common sense
Authored by: Anonymous on Thursday, June 17 2004 @ 08:56 AM EDT
Look folks, we aren't buying the Open Source community any credibility by
clinging to this notion that somehow Novell didn't transfer the copyright to

Ask yourself this - if Novell didn't sell ownership of the copyright to SCO,
then what exactly was SCO buying? What did SCO spend all that money on?

It is time for us to cede this point to SCO and move on.

I really hope this post doesn't get deleted, as I believe in Linux and only want
the worst for SCO.

[ Reply to This | # ]

Can this be sent to the relevant attorneys?
Authored by: Anonymous on Thursday, June 17 2004 @ 09:03 AM EDT
While it appears that some of the judges and attorneys involved in all of this
do use the internet now and then to research, can this bit of digging be sent to
the relevant attorneys involved in the Novell vs. SCO case? Or, would that be

[ Reply to This | # ]

not DCE, maybe a modified xdm renamed?
Authored by: jesse on Thursday, June 17 2004 @ 09:35 AM EDT
Although oldSCO appears to be claiming a copyright going back to 1976, there's only one package with a COPYRIGHT file that mentions 1976. Strangely, it's for the CDE login manager, which I believe didn't exist back in 1976.

It is true that CDE did not exist in 1976, but it is also true that most of the login manager DID exist as part of the MIT xdm display manager. I suspect this MAY refer to some old code that oldSCO had added to xdm for their purposes. Perhaps early NDIS support for instance.

[ Reply to This | # ]

How could Novell have sold other people's copyrights?
Authored by: Anonymous on Thursday, June 17 2004 @ 09:45 AM EDT
First I'd like to congratulate Dr Stupid on an interesting and well-written

Couple of other points occur to me:

1. SCO's position appears to be that they acquired the copyrights to the whole
of UNIX by their interpretation of the APA and Amendment 2.

Novell the APA and Amendment 2 are agreements between Novell and Santa Cruz

What I think is a major hole in this theory, is there are numerous other
copyright holders in UNIX (as cited by Dr Stupid), so even if Novell did assign
some Novell/USL/AT&T copyrights (I don't think they did but arguendo), --
Novell could *not* have assigned copyrights that did not belong to Novell, for

Computer Associates International, Inc.
Edison Design Group, Inc.
Eric P. Allman
Hewlett-Packard Company
Hitachi, Ltd.
Intel Corporation
International Business Machines Corporation
Massachusetts Institute of Technology
Microsoft Corporation
The Regents of the University of California
Sun Microsystems, Inc.
The Open Group (formerly OSF)
Compaq Computer Corporation
Digital Equipment Corporation

In short:
Unless SCO can show copyright transfers (section 204a writings) from each of
these to Caldera via Novell and Santa Cruz -- then SCO simply can never show
their copyright claim to the whole of UNIX is valid.

2. Based on the fact that UNIX was assembled from contributions from multiple
copyright holders (including those listed above - and it was established in the
BSD trial public records that AT&T took a license from University of
California for BSD stuff) -- it rather sinks SCO's latest theory that UNIX can't
be split into parts (see SCO's memo on discovery for this rather bizarre claim)

[ Reply to This | # ]

Is there a precedent of presumption here?
Authored by: Anonymous on Thursday, June 17 2004 @ 10:02 AM EDT
I am not a lawyer and I was wondering about case history here.

1) Is there a presumption towards the original owner of the copyright? In other
words if there is a dispute as to who owns the copyright, the burden of proof
should rest with the party that claims to now own the copyright. If his
evidence is not strong enough to prove he owns the copyright, the previous owner
retains ownership. Is this correct?

2) Isn't rare for there to be a copyright transfer dispute case? Not a case
about who had the copyright first but a dispute about whether a copyright was
transfered or not.

3) What is "registering" a copyright? Why can SCO just automatically
get this registration. Was Novell suppose to contest the registration? Can the
contest it? What advantage does "registering" you copyright have?

Just some questions. If no. 1 is true, it definitely strengthens Novells case.

[ Reply to This | # ]

Another piece of the puzzle:
Authored by: Anonymous on Thursday, June 17 2004 @ 10:07 AM EDT
To expand upon "A 1996 Clarification of What SCO Purchased from Novell":

At Subject: UnixWare/OpenServer Watch #11

There is the reference to:

    "...some of the writing might have confused some readers into thinking that SCO had taken the reins of the entire UNIX System business, and not just the UNIX System licensing and UnixWare businesses SCO purchased from Novell..."

    "...Usually, X/Open's guide to the use of its trade marks isright next to me when I create UOW..."

The following documents when "X/Open" metamorphosed into the current Open Group as a merger between "...X/Open Company, Ltd. and the Open Software Foundation (OSF)..."


    "...SAN FRANCISCO, CA (February 14, 1996) -- X/Open Company, Ltd. and the Open Software Foundation (OSF), the two leading consortia for the advancement of open systems, today announced their consolidation into a new, more powerful worldwide organization, known as The Open Group. The new entity has been formed to strengthen and streamline the entire open systems process, including adoption of open systems specifications, development of specification-compliant technologies and promotion of their use in the global enterprise computing marketplace..."

Currently, The Open Group:

    "...holds the definition of what a UNIX system is and its associated trademark in trust for the industry.

    In 1994 Novell (who had acquired the UNIX systems business of AT&T/USL) decided to get out of that business. Rather than sell the business as a single entity, Novell transferred the rights to the UNIX trademark and the specification (that subsequently became the Single UNIX Specification) to The Open Group (at the time X/Open Company). Subsequently, it sold the source code and the product implementation (UNIXWARE) to SCO. The Open Group also owns the trademark UNIXWARE, transferred to them from SCO more recently..."

See: tml

There is a UNIX history at:

http://www.unix.o rg/what_is_unix/history_timeline.html


I immediately archive every Groklaw page to which I make a comment, for the record.

[ Reply to This | # ]

Unixware updates and copyright notices
Authored by: Anonymous on Thursday, June 17 2004 @ 10:26 AM EDT
Now that they have "released" a new update to unixware, I
wonder how SCO has changed the copyright notices within
it. Actually, claiming other people's copyrights in a
work seems to at least border on the idea of "misuse of
copyright", a federal doctorine that allows copyright to
be terminated by the court. It would be interesting to
see it applied to one of the active cases...

[ Reply to This | # ]

But where is Old SCO's transfer to New Sco?
Authored by: JRinWV on Thursday, June 17 2004 @ 10:37 AM EDT
I don't recall seeing a copy of the contract between the4 old SCO and The SCO
Group transferring copyrights. Does Groklaw have that document yet?

[ Reply to This | # ]

This is not unique for UnixWare..
Authored by: Anonymous on Thursday, June 17 2004 @ 10:42 AM EDT
I'd just like to note that this adding-copyright-statments stuff is not unique
to UnixWare..

For instance the Digital Unix stdio.h includes the following:

Copyright (c) Digital Equipment Corporation, 1991, 1997
Copyright International Business Machines Corp. 1985, 1988

[ Reply to This | # ]

Nice article - but ...
Authored by: n0ano on Thursday, June 17 2004 @ 10:50 AM EDT
First I'd like to compliment you on a well researched and presented article,
great effort on a good job.

Unfortunately, the reality is that there is no hidden meaning in Unix copyright
notices. In general, programmers (at least myself and others who started on
Unix at Bell Labs) hate dealing with copyright notices. All they are are drudge
work that are just there to make the lawyers happy[1]. I can well believe that
oldSCO programmers would make changes to files and leave the old Novell
copyright in place, even though technically a new oldSCO copyright should have
been added.

Note that no individual names are placed on any of the Unix copyrights, only
company names. Us programmers never had a sense of ownership of the actual
files so we didn't overly care about the copyright notice. Make the changes,
test the code, check it in and move on to the next problem. The only time we
worried about copyrights was when a manager got religion and forced everyone to
take time out to bring copyright notices up to date.

Also, very few programmers know (or care) about the intricacies of what
copyright notices should, or should not, contain. I know that, personally, I
would NEVER remove someone elses copyright, that would be tantamount to stealing
their code. I would, and have, added a new copyright on top of an old one,
working on the assumption that the new copyright supercedes the old one, but I
always leave the old on in place.

I agree with you that Novel, more through luck than intent, still retains the
copyrights to Unix and only sold oldSCO the distributor channel, that is what
oldSCO spent all its money on, but I have to say that the presence of Novel
copyright notices in oldSCO releases provides no evidence for, or against, this

[1] Who knew that the lawyers were right and we really should have been
concerned about copyrights. At the time it seemed like stupid busy work but now
the whole issue has come back to haunt the entire industry - go figure.

"Censeo Toto nos in Kansa esse decisse." - D. Gale

[ Reply to This | # ]

What's confusing here is that it looks like SCO has source code but no copyright
Authored by: BubbaCode on Thursday, June 17 2004 @ 10:56 AM EDT
Normally if you OWN the source, not just have a copy for your use, you OWN the
copyright. The only way to explain that Novell owns the copyright is that they
also own the source code. I believe they do.

SCO has the right to use the source code as they wish, sell UNIXWARE etc., but
ultimately the code is Novells. They could even sell the code to someone else
if the agrements between them and SCO were not exclusive (which I believe they

This is VERY common in the SW development community believe it or not. Many
companies will sell you a source code included license. You can use and modifiy
the source code to your needs, but the original souce code copyright is retained
by the company that sold you the source code and they can sell the code to as
many people as they wish. Usually the purchaser is encouraged to by a
maintainence license that gives them updates as the original source code is
changed. This is common for source code included embedded system kernels and

For example we could buy an OS that includes the source code. We may latter buy
from the same company or someone else an add on library that adds a feature we
need. Our company can change the code and sell products based on this code.
Depending upon the agreements we would either pay a one time up front fee (like
SCO did) or pay royalities. There is usually limitations to the license of the
source code. For example we are usually limited to only one product per
licensed copy.

This is basically what SCO excuted when it bought UNIX from Novell. The
exception to this example is that they also bought some exclusive rights where
Novell was prevented from doing certain things with thier copyrighted code.
This should NOT be confused with a transfer of copyright.

BTW, by this standard Novell does not own the changes SCO did to UNIX on their
own, just the original code sent to SCO and any changes Novell has made.

[ Reply to This | # ]

Nice work, Stupid. Er ;)
Authored by: Anonymous on Thursday, June 17 2004 @ 10:56 AM EDT
This nice write up puts something Darl keeps saying into perspective for me...

"What did we pay $100+ millions for?"

Hmmm, what did Sun pay $81 million dollars for? Sun was ripped off! For a bit
more they could own it all! Instead they just bought a simple license to

[ Reply to This | # ]

Ex Caldera CEO Ransom Love in 2000: "We *don't* own the IP to open server"
Authored by: NZheretic on Thursday, June 17 2004 @ 11:43 AM EDT
Your going to need to use the old realplayer to view the stream,
realplay rtsp://

From Technetcast LinuxWorld 2000:Ransom Love Keynote

Linux and the Chasm: Where Are We? Caldera President Ransom Love explains what it will take for Linux to be adopted accross the entire landscape... Also: SCO acquisition... COSMOS, a workstation management system...
At the end of the speech Ransom Love clearly states:
There is one element of Open Server that is not coming over, we don't own the IP , we just own all the rights for distribution, ongoing development for the Open Server and that has to do with tax and other considerations

[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: Anonymous on Thursday, June 17 2004 @ 11:45 AM EDT
Very nice analysis, and very helpful! In absence of clarifying language in a
document somewhere, and with the APA being the hellishly written document that
it is, the court is going to have to look at what the parties were doing at the
time of the transfer to determine the intent. SCO and Novell can say whatever
they like, it likely doesn't matter to the judge other than to determine that
they disagree. The behavior of both parties around the time of the APA and the
ammendments will, IMHO, likely determine how the court will rule on the intent
of the APA.

Bottom line is with SCO leaving Novell copyrights on UnixWare until well after
ammendment 2 time, and the absence of any document stating flatly that Novell
transfers the copyrights (and which ones specifically) to SCO, SCO can't prove
ownership of the original UNIX SysV kernel code (which is what this whole
argument is about). If they don't own the copyright, they can't sue anyone for
infringing the SysV kernel code, only the copyright holder can do that.

However there's a larger issue at hand than just the copyrights with respect to
the APA that Dr. Stupid is addressing here, which is the overall intent of the
APA with regards to what rights Novell retains and what rights SCO gets. Beyond
the issue of asset transfer, there's a larger issue (at least in scope of what
it will affect in SCO's litigation) concerning what Novell has the right to do
under the agreement. Novell would seem to claim to retain the rights to force
SCO to waive claims against UNIX SVRX licensees (the targets of all current
litigation mind you, even Novell although they don't pay SCO for it). That being
said, if the judge rules wholely on the APA's meaning and intent (which would be
a good idea in terms of judicial efficency and the fact that both parties are
facing each other over APA issues already), then SCO will sink or swim as a
whole on this battlefield.

What I mean is it would be a great service to everyone if Judge Kimball would
clarify the meaning of the entire document, including defining what rights each
party retained or were granted in the document as ammended (it is ammendment 2
afterall that lawyers should be flogged for writing it's so ambiguous). A great
service, no matter what the outcome (although I suspect things are most likely
as Novell says they are). At the very least, any confusion about what SCO's
current rights are would be defined so as not to be an issue of contention in
any other court case, and that at least speeds things along to conclusion.

SCO doesn't want things to happen fast so I would guess they would vigorously
oppose any attempt by Kimball to rule on the above stated merits of the APA,
although they likely need a court victory somewhere to convince anyone to take
an IP license. The flip side is they lose in this venue and the rest of their
litigation fleet gets dragged down with this ship, leaving them the target of
numerous lawsuits.

After all, if Novell can waive SCO's claims, then the IBM, AutoZone, and DC
cases are done where they stand, since SCO would have no legal basis to bring
suit (no rights to be infringed with respect to UNIX SysV) against any of them
due to Novell's previously publicly announced directives to SCO to waive the
claims against all three entities.

[ Reply to This | # ]

(c) 1976?
Authored by: Anonymous on Thursday, June 17 2004 @ 11:51 AM EDT
I remaining intrigued by the references to Copyright © 1976-1998 The Santa Cruz Operation, Inc.

If you take a look at almost any UNIX timeline/history, 1976 is just about back at the dawn of time, as it were; May 1976 marks the UNIX Time Sharing System Sixth Edition (V6).

I'm wondering if The Santa Cruz Operation copied AT&T's entire copyright line intact, changed it to read "The Santa Cruz Operation", and merely did an inconsistent job of adding that line back into specific files.

Thus the presence or absence of that specific copyright line in any given file is of little significance.


I immediately archive every Groklaw page to which I make a comment, for the record.

[ Reply to This | # ]

OT: SCO UnixWare drops gcc, LKP
Authored by: Anonymous on Thursday, June 17 2004 @ 12:43 PM EDT
SCO has removed the oss from UnixWare for the new release


[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: Anonymous on Thursday, June 17 2004 @ 12:59 PM EDT
In the APA there is another rather strong indication that Novell retainde the
copyrights and so forth to sysvr6.

APA 4.16b (parts):
"Buyer shall not, and shall have no right to, enter into future licenses or
amendments of the SVRX Licenses, except as may be incidentally involved through
its rights to sell and license the Assets or the Merged Product (as such term is
defined in the proposed Operating Agreement, attached hereto as Exhibit 5.1(c))
or future versions thereof of the Merged Product."

So oldSCO were allowed to sell licenses for Unixware, the merged product etc,
but were NOT allowed to enter into future licenses and amendments of sysv

Anyone care to comment?

[ Reply to This | # ]

OT: And another thing - C++
Authored by: ihawk on Thursday, June 17 2004 @ 01:54 PM EDT

Remember Darls mouthing about ~"We also own C++. We get a lot of money for C++"? (that symbolic construct of '~"' there is my attempt at saying this is an approximate quote. :-))

O'Reilly has developed a poster of the history of programming languages. It has C++ as a sort of child of K&R C and Simula, beginning in 1980 as "C with Classes". By 1998 (that would be after newSCO acquired whatever it was they may or may not have acquired, right?), C++ is listed as ANSI/ISO which sort of implies that it is a standard.

Okay, so that doesn't mean that newSCO couldn't have an implementation of C++ that someone has been stupid enough to pay them for, but it does kind of tend to cast doubt on the "We own C++" thing.

BUD: Beer, Uncertainty and Doubt. In that order.

[ Reply to This | # ]

Authored by: Anonymous on Thursday, June 17 2004 @ 01:56 PM EDT
What can we say?

In your face, Darl!
DrStupid just ate your lunch, and you might as well pack your pencil and vacate
your corner office.

[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: Anonymous on Thursday, June 17 2004 @ 02:05 PM EDT
There is a basic principle in contract law that has not, up to this point, been
fully expounded here at Groklaw, and it SINKS David Bois' arguments regarding
the line "if we didn't get copyrights what did we pay for?"

A Contract law bedrock principle is that "the law will not inquire into the
adequacy of consideration" in evaluating the formation or the terms of a

In other words, the only proof needed re proof of formation is that SOME
consideration is passed. The courts take no notice as to the amount, whether
"too little" or "too much". Once the threshold analysis that
'some' consideration was exchanged, the the contract exists. The CONTENT of the
contract is then defined by "the four corners of the document" unless
the document is DIRECTLY ambiguous, i.e., if it contains directly contradicting

To argue that the contract "must" mean x because I paid y money is an
absolute NON-STARTER in pure legal terms, and SCO will have GREAT DIFFICULTY
arguing that with a straight face before a judge.

[ Reply to This | # ]

Is this why Darl, Blake, Chris, Ralph et al have to lie?
Authored by: Anonymous on Thursday, June 17 2004 @ 02:10 PM EDT
You said:

"Overall, oldSCO's stance looks like an assertion of collective copyright -
that oldSCO had the copyright on the specific compilation and integration of
programs and code that made up UnixWare, and could thus pursue those who copied
UnixWare as a whole..."

This may be the reason Darl et al have to lie about millions of lines of code:
if it isn't a whole, they don't have a leg to stand on - and it is now well
proven, documented fact that there is no improperly copied code in Linux, let
alone millions of lines.

But your article seems to strengthen the various statements we've heard that SCO
and SCOG may not have clean hands from the other direction. LKP and other
projects, when combined with the blatant abuse of copyrights you point out,
verge on the criminal. Combine with perjury and lying to various agencies and
authorities, and perhaps the line is crossed?

[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: PhilTR on Thursday, June 17 2004 @ 02:16 PM EDT
One must also remember just how much Novell paid to USL for UNIX. As I recall it
was $900,000,000.00. And just exactly how much did SCO pay for its "right
to use" UNIX? Also, users have been licensing software to support and
operate their business and nary a copyright ownership ever changed hands. philtr

[ Reply to This | # ]

Where's the money?
Authored by: clacour on Thursday, June 17 2004 @ 02:39 PM EDT
One strong argument in one direction or the other would be how much money
changed hands when Amendment 2 was made.

The original agreement made it crystal clear that copyrights were NOT part of
the deal.

My reading of the clause in amendment 2 was that Novell was saying "No,
we're not screwing you by selling you the right to license Unixware, but then
holding the copyrights hostage. If you need one of the copyrights in order to
exercise the rights you DID buy, we'll give it to you."

Under that interpretation, there would probably have been no payment by oldSCO
to Novell - it was simply a clarification and reassurance that there would be no
nasty surprises on down the line.

If, on the other hand, oldSCO paid an additional $200 or $300 million as part of
amendment 2, they've got a strong argument that yes, oldSCO DID purchase the
copyrights with that agreement.

If the amount of money was intermediate, one would have to judge what the rest
of the agreement was worth. If there wasn't a payment at least $150 million more
than whatever else that was in amendment 2 was worth, newSCO is on very shaky

Contract law requires there be value for both sides for it to be a valid
contract. A contract change that results in a couple hundred million more value
for SCO and nothing for Novell is not going to be considered legitimate.

Does anyone know how much money (if any) changed hands when Amendment 2 was

[ Reply to This | # ]

Something I've thought for a while
Authored by: Anonymous on Thursday, June 17 2004 @ 03:19 PM EDT
These glimpses into the progression of notices during & following the
initial APA supports something I've been thinking all along:

The amendment did transfer *some* copyrights, because there was clearly a need
for some, i.e. the manuals. oldSCO would have to distribute documentation &
manuals, and they would need some copyrights to do so.

Somewhere in all the volumes of data posted here I remember seeing a list of
manual copyrights submitted (I think) by newSCO regarding perhaps the AZ case.

Does anyone know where that list came from and when? I'm thinking that maybe
there WAS a list of what was required (and transferred) and that list was the
extent of it.

newSCO is relying on the fact that Novell obviously wasn't keeping the best
records (since they didn't even remember the 2nd amendment at first) and hoping
they've lost that list altogether.

My logic is that since the APA excluded all copyrights, but that some were
clearly needed, and since the amendment doesn't specifically say which ones,
that they must have been detailed somewhere separately.

That would also explain some of the overlapping dates in the published notices,
since Novell would also need the same manuals & documantation for the
licensees they still retained rights over, therefore, while they would assign
some rights to SCO, they would also retain some for themselves.

Somewhat logical, or is my tinfoil hat starting to meltdown?

[ Reply to This | # ]

Could Novell have sold the copyrights?
Authored by: Anonymous on Thursday, June 17 2004 @ 05:07 PM EDT
If Novell had already entered into any licence agreements with any other party
(Sun?) they may not have been in any position to sell the copyrights. After
selling someone a licence to do something with their copyright material they
could hardly sell the copyrights to someone else without making the new
copyright holder honour all the existing licence agreements. Such an arrangement
would surely produce a lot of paperwork saying who the existing licensees were
and what the licences were for. No one has suggested that any such documentation
exists. To lose your life assurance policy is unfortunate, to lose your cheque
book and car keys as well looks like carelessness.

[ Reply to This | # ]

Some early SCO(1996ish) Unixware Faqs and how to's
Authored by: SteveS on Thursday, June 17 2004 @ 05:15 PM EDT
History! I love history,

Well sort of, I ran across this stuff while trying to find a ref to uww #10. Thought I would drop a link here for safe-keeping...

Nice history of Unixware.

Steve S.

[ Reply to This | # ]

More Fodder, but a shade OT
Authored by: ihawk on Thursday, June 17 2004 @ 05:26 PM EDT
I'm beginning to get real hung up on the question of what oldSCO provided to Caldera. That's as much of a question as what Novell sold to oldSCO, because oldSCO could not have sold what they didn't have.

So I found this while poking around - it's an archived Infoworld article from 2000 talking about the SCO -> Caldera sale of Unix and UnixWare technology.

One very interesting quote in the article from a D.H. Brown analyst named Tony Iams:

"Caldera was a pure Linux distributor, but now they have all this closed-source Unix from SCO," Iams explained. "They can't just open up the closed code ... they can't afford to open it up. What is there left to sell if everything is open source? So they're thinking of making it more open, but slightly restricted." Iams said the Unix code that Caldera received contains licensed technologies from previous suppliers, namely AT&T and Novell, which create roadblocks in opening up the source code.

That seems to convey the idea Dr. Stupid is supporting here, that Novell (along with AT&T) maintained it's copyrights in the Unix codebase.

(A slightly interesting side note to this article is that it mentions Doug Michels as being with Caldera - in fact he was the president of oldSCO at the time. Inaccuracy in reporting is nothing new.)

There's another interesting article at Forbes (yeah, I know, boo hiss) that has an enlightening quote or two:

Caldera will form a holding company, Caldera Inc., which will hold the assets purchased from SCO, including its employees and exclusive rights to distribute SCO's OpenServer and UnixWare operating systems. In the first jab at consolidation, Caldera will merge Linux and Unix development and support programs, making them more compatible.

That one paragraph has all kinds of tantalizing implications (the reference to purchasing SCO employees notwithstanding ;-)) - for instance the "exclusive rights to distribute" wouldn't include copyrights, and the reference to merging the development of Linux and Unix which might tend technology from one platform to the other, no?

I'm still looking for more solid references to what was sold. The original deal apparently didn't involve OpenServer because it was no longer under development, but it confused customers, so Caldera and oldSCO changed the deal in 2001 to include OpenServer in the asset transfer. There are some other hints, but nothing solid. It really bothers me that newSCO apparently lost (!?!) the agreement that conveyed these assets and we haven't heard about the specific details of it from Taratella or anyone. I would assume that IBM and/or Novell have asked Tarantella to provide this information, but I haven't heard about it showing up yet.

Inquiring minds want to know, dagnabit!

[ Reply to This | # ]

Failure to market aggressively gets SCO's license yanked?
Authored by: edumarest on Thursday, June 17 2004 @ 06:36 PM EDT
This has to be the longest lists of posts I have seen yet and on a dialup at
28.8 there is much that I can overlook. My apologies if this has been mentioned
elsewhere. One thing I did notice was:

"Amendment 2 states that

"Novell may execute a [System V royalty] buy-out ... without any approval
or involvement of SCO, ... if: (i) SCO ceases to actively and aggressively
market SCO's UNIX platforms.."

That seems to say that Novell could seek to generate revenue from royalty
buy-outs if oldSCO lost interest in UNIX. This reads not as a right granted to
Novell, but as a right held but restrained as long as oldSCO "actively and
aggressively" marketed UNIX."

So if oldSCO really owned UNIX then why did Novell have the right to take it
back? That is a pretty powerful argument right there.

Dr. Stupid, your fact-finding was excellent.

...if you cannot measure it then you cannot troubleshoot it, you can only
SuSE 9.0 on hp pavilion ze 4560us

[ Reply to This | # ]

Notice this Notice? ~ by Dr Stupid
Authored by: Anonymous on Thursday, June 17 2004 @ 07:37 PM EDT
Back in the early 90's when FPS games started becoming popular I adopted the
name GoMMeR. Largely because of GoMMeR from the Andy Griffith show. My thinking
was that if you frag someone named GoMMeR - no big deal. No ego boost, notta.
After all, the guy thinks of himself as a goof anyway.

But if a guy named GoMMeR repeatedly frags you, ohhh that's sad!

I can think of a more fitting name to pwn SCO then Dr. Stupid. Well, maybe MBA

[ Reply to This | # ]

Here's An Amusing Thought
Authored by: tredman on Thursday, June 17 2004 @ 07:50 PM EDT
Considering for a moment that TSG has reseller and distribution rights to UNIX,
instead of wholesale ownership of the operating system. Since the APA with
Novell, they've been revising and updating UnixWare and OpenServer off and on
over the years.

If they didn't purchase the copyrights from Novell, and nobody with knowledge of
this case can honestly say that it seems they did, how fast do you think TSG
would have to spin their whole "derivative works" theory, since they
did exactly the same thing that IBM, Sun, HP, Digital and every other UNIX
vendor does.

The image in my head is of Darl McBride stuttering like Mel Tillis.


[ Reply to This | # ]

OT: Daimler Chrysler?
Authored by: Anonymous on Thursday, June 17 2004 @ 08:11 PM EDT
Okay, this is fun, but I'm ready for the next chapter of SCOmedy

Anybody seen SCO's opposition to DC's motion for Summary Disposition?

[ Reply to This | # ]

2 press releases to enjoy
Authored by: Anonymous on Friday, June 18 2004 @ 12:39 AM EDT
You can find many original SCO press releases on comp.unix.sco.announce
(archived for example in google groups)

Here's a couple of good ones:



December 3, 1996 <-- supposedly after the date when Novell lost any interest
in UNIX according to podSCO



April 25, 1996 <--- supposedly after the date when SCO acquired copyrights

Including the interesting phrase "We saw product margin
improvements mainly due to the UNIX technology acquisition from
Novell and the recent buy-out of the TCP/IP technology which
resulted in the highest gross margin level in the company's

Now the question is what were they doing buying-out (i.e. getting a royalty free
license) TCP/IP technology if they supposedly already owned the copyrights to
all UNIX already?

You can also find Ransom Love's press release about Caldera acquiring two
divisions from Original SCO in the same group's archive. This one is too
hilarious in retrospect. According to this press release, Caldera bought these
two divisions from SCO inorder to make Linux the standard and perhaps replace

[ Reply to This | # ]

Scrivner's error
Authored by: AllParadox on Friday, June 18 2004 @ 01:02 AM EDT
This is really an equitable solution. Sometimes, a court will put on its
"equity" hat and will rewrite a contract. This is done because
something critical was accidentally left out. The essence of the solution is
that there was an accident. If there were 27 versions of the document, and
versions through 26 had a one-page clause, but v27 was missing the page and the
clause, and the page number was also missing, then there would be evidence for
"scrivner error". (i.e. they lost the page with the clause).

Courts are extremely reluctant to find "scrivner error". They will
not make the finding unless there is a clear error and evidence of an accident.

Courts are even more reluctant to find "scrivener error" when it is a
major portion of the contract. It is one thing to accidentally misspell
something or leave it off a long list. It is quite another to forget the list
entirely, especially when the whole contract is supposed to be about the list.

Finally, courts consider the amount of scrutiny a contract has received. A
quickie two pager between two small businesses will be amended much more quickly
than a hundred-million dollar contract between two major corporations, reviewed
and endorsed! by their legal representatives.

Suggesting that both corporations, and both corporations' legal teams,
overlooked listing copyrights when they included everything else possibly
relating to copyrights or intellectual property, is just silly.

All is paradox: I no longer practice law, so this is just another layman's
opinion. For a Real Legal Opinion, buy one from a licensed Attorney

[ Reply to This | # ]

Microsoft and Old SCO
Authored by: Anonymous on Friday, June 18 2004 @ 01:07 AM EDT
New SCO contends they got all the copyrights to Unix, via Old SCO, from Novell
in 1995 or 1996

If that is the case, why did old SCO have a license from Microsoft to include
certain Microsoft copyrighted code - and go to the EU to force a revision of the
license in 1997

[ Reply to This | # ]

changes in UnixWare 7.1.4 disty
Authored by: Anonymous on Friday, June 18 2004 @ 03:39 AM EDT
Did anyone else see the news of this? It's visible at

Is this happening as a reaction to
SCO being in default of the GPL, I wonder? Or is it part of some move towards
carrying zero OSS content, in reaction to rebuttals of their position which
point out their continuing use of OSS?

[ Reply to This | # ]

Smoking gun? TTLA and SCOX
Authored by: mk270 on Friday, June 18 2004 @ 05:16 AM EDT
The Yahoo finance board on SCOX had some good articles about the transfers
between the companies now known as Tarantella and The SCO Group. The poster was
analysing TTLA's filings with the SEC, and trying to work out whether SCO was
paying TTLA licensing revenue for onward payment to Novell.

The posts are numbered 128753, 128767, 128805, 128806 and 128808.

Go to and type these numbers
into the appropriate little box.

[ Reply to This | # ]

Semi-OT: How could copyright ownership ever be required?
Authored by: LouS on Friday, June 18 2004 @ 07:10 AM EDT
The ammended APA speaks of copyrights "required by [oldSCO] in order to exercise its rights" in the technology it was purchasing. But the copyright owner can give someone else any of the rights the owner has without transferring the copyright ownership per se. So how could copyright ownership ever be required by oldSCO in order to exercise its rights?

[ Reply to This | # ]

OT SVRX copyright ownership does not matter
Authored by: QTlurker on Friday, June 18 2004 @ 08:48 AM EDT
Once again.

I think we are forgetting the main issue. Linux and the GNU utilities, together,
are an independent implementation of Unix.

A lot of thought (technical and legal), talent, and effort was expended by a
large cast of characters to ensure that GNU and Linux can not be considered
derivative works of AT&T DNA Unix.

It does not matter who owns the AT&T legacy copyrights -- SCO, Canopy,
Novell, AT&T, Intel, or the public. Linux and GNU doesn't infringe!

Fortunately, IBM and Redhat seem to understand.

[ Reply to This | # ]

Actually, the SVRX copyrights matter a great deal...
Authored by: Anonymous on Friday, June 18 2004 @ 09:31 AM EDT
They are the basis by which SCO is currently pursuing IBM in court. SCO's
current claim against IBM is mainly a copyright infringement case, since they
have dropped their trade secret claim. That infringement is based upon IBM's
refusal to cease shipping AIX after SCO allegedly terminated their license.

In order to claim copyright infringment in this way, you must have rights to be
infringed upon in the 1st place. If it were proven in court that SCO didn't own
the SVRX copyrights as they claim to, their case against IBM is largely over.

So the copyrights matter a great deal. SCO v IBM has nothing to do with whether
Linux infringes UNIX SvsV copyrights, only whether SCO has any copyrights for
IBM to infringe by continuing to ship AIX allegedly without a license to do so.

[ Reply to This | # ]

OT - Is SCO acutually suing anyone over general infringement in Linux?
Authored by: Anonymous on Friday, June 18 2004 @ 09:41 AM EDT
After going over the different cases, it would seem that despite SCO's very
public claims of Linux massively infringing SCO's rights, they are in fact not
suing anyone over this issue.

1) SCO v IBM:
SCO claims copyright infringement, but this is born out of IBM's continuing to
ship AIX after SCO purportedly terminated their license. Has nothing to do with
Linux infringement (no trade secret or copyright claims there)

2) SCO v Novell:
Contract dispute. Hopefully will clarify the APA (or at least some of it)

3) SCO v AutoZone:
Infringment by use of Linux, but claims it arises because AutoZone or IBM must
have copied libraries directly to Linux for their migration to have worked.
Nothing to do with whether Linux as an OS infringes.

4) SCO v Dhaimler Chrysler:
Contract dispute (or stupidity, take your pick). Argument over whether DC must
comply with an audit request from SCO when they are not a current licensee.
Again, nothing to do with Linux at all.

So they keep running around telling everyone that Linux infringes and they have
erected this great smokescreen of lawsuits that look like cases against people
over Linux infringement, but in fact they are suing NOONE over Linux
infringement. They might have to prove it in the IBM case to even prove they had
a cause to terminate IBM's license, but I don't see any legal claim against IBM
for the alleged contributions.

No wonder no one will buy a license, SCO doesn't seem to want to get into the
issue directly in court. I wonder why......... (massive sarcasm alert)

[ Reply to This | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )