Perhaps SCO feels it can't find sufficient living people to agree with their interpretation of the license, so as to counter the united army of IBM supporters, so they instead dig decades back to find a letter on Otis Wilson's letterhead -- but signed for him by someone else -- to say, in effect, that Mr. Wilson's declaration, also in support of IBM, is somehow rebutted by this old letter to DEC.

Mr. Wilson, who was "head of the group responsible for licensing the UNIX System V operating system worldwide" at AT&T and later USL, testified:

"AT&T and USL did not intend to assert ownership or control over modifications and derivative works prepared by our licensees, except to the extent of the original UNIX System V source code included in such modifications and derivative works. Although the UNIX System V source contained in a modification or derivative work continued to be owned by AT&T or USL, the code developed by or for the licensee remained the property of the licensee, and could therefore be used, exported, disclosed or transferred freely by the licensee."

He said that one reason why IBM's contract was not intended to control their work was that IBM wouldn't have agreed to sign the contract if it had such terms, and, secondly, AT&T was under antitrust scrutiny at the time, and the company was extremely sensitive about potential anticompetitive actions, or even the appearance of them, at the time the license was negotiated. For that reason, he testified, *when licensees asked for* reassurance that the terms were not seeking such control, they happily provided written or oral clarification, in IBM's case written, in the Side Letter.

What SCO would dearly like to prove with the DEC letter is that AT&T did so try to protect methods and concepts. Maybe back in 1984 they may have tried with DEC in this one particular deal, but what does that matter now? We've seen what USL thought it could still protect in 1994, in the USL-Regents Settlement Agreement that recently came to light, 26 piddly "Restricted Files", and in my mind, it makes all these discussions about 1984 and methods and concepts so totally over. And anyway, IBM has the Side Letter.

Not that a passing reference to 1984 doesn't feel somehow appropriate when discussing SCO's stated position. Here is what SCO said about this letter:

"Third, IBM’s declarants now deny that AT&T sought to protect the methods and concepts embodied in UNIX System V, again in the face of plain contractual language to the contrary. Mr. Wilson’s declaration now states that, at least as of February 1985, AT&T did not seek to protect the methods and concepts embodied in UNIX System V:

'In fact, we were not aware of any particular "methods or concepts" that needed to be protected. We made up this phrase as sort of a general catch-all. We were willing to delete the reference to methods and concepts in the IBM agreement because we were not aware of any UNIX System V methods or concepts that required protection. The fact that we had distributed the UNIX System V source code so broadly that the internal structure of the UNIX system V operating system was well known in the academic community and by computer programmers generally.' Wilson Decl. (Dec. 2003) ¶ 14 (Exh. 21).'

"See also Frasure Decl. ¶ 12 (Exh. 18) (claiming that the licensing agreement provisions were not intended 'to restrict our licensees’ use export, disclosure or transfer of anything besides the licensed UNIX System V source code and related materials').

"But just prior to and well after the IBM Side Letter was signed in February 1985, Mr. Wilson (on behalf of AT&T) repeatedly and expressly sought to protect UNIX 'methods and concepts' pursuant to AT&T’s license agreements. For example:

• In his letter to Digital Equipment Corporation (“DEC”) dated Sept. 5, 1984, Mr. Wilson wrote: “Regarding Section 2.01 of the referenced Software Agreement, you do not have all right, title and interest in derivative works based on a SOFTWARE PRODUCT if such derivative works include any of our code or embody any of the methods and concepts used in a SOFTWARE PRODUCT.” Exh. 22 at 2 (emphasis added).

• In proposed (and later executed) letter agreements dated June 18, 1990, June 29, 1990, and August 3, 1990, Mr. Wilson stated to other licensees: “You will not provide access to any copy of the source code of the SOFTWARE PRODUCT (including methods and concepts contained therein), in whole or in part, to anyone other than your organization’s employees who have a need to know.” Exhs. 23, 24 & 25 (emphasis added).

"Such statements plainly belie Mr. Wilson’s current claims that by the time of the IBM Side Letter in 1985, 'we were not aware of any UNIX System V methods or concepts that required protection' and that AT&T 'had distributed the UNIX System V source code so broadly' that no protection for methods and concepts could obtain."

What Mr. Wilson said was that when IBM pressed the issue, AT&T was willing to strike that methods and concepts language, because it wasn't meaningful any more. That doesn't speak to 1984 with DEC, or, for that matter, what he might not bother to strike from their boilerplate language in a license with someone who didn't press the issue the way IBM did. One thing I understood from the USL-Regents settlement is that a lot of folks may have been paying for pretty much thin air for a long time. No wonder they wanted the agreement to be kept a secret.

There are a number of reasons why this exhibit might not be persuasive, even if we hadn't seen the USL-Regents settlement. The date matters, for starters. As you will recall, AT&T's Wilson testified in his deposition in 1992 in the BSDi case that everything changed some time in 1984. IBM's original commercial license was executed in 1985, after the change. There was a liberalization as a general trend, in harmony with what Mr. Wilson described as his company's desire to spread Unix widely in the educational and commercial sphere, with the result that, according to Mr. Wilson, very little, if any, of Unix is confidential any more. His testimony certainly matches what we saw in the unveiled USL-Regents Settlement Agreement, where USL was left with only a handful of files that no one needed any more and could live without happily ever since.

SCO, however, has dug among the cobwebs and would like to point out that in this particular letter, AT&T was trying to keep certain code confidential. All right, but that is decades ago, and while SCO would like to turn back the hands of time, the world in fact moved on, and so did Unix. There has been a lot of water under the bridge since 1984, yet SCO creepily clings to the past like the eery Miss Havisham in her wedding dress, sitting at a table of stale cake in "Great Expectations".

This is a letter agreement where the parties are agreeing to some things and agreeing that they'll work out some other things later, and it's some kind of unusual arrangement for work on a particular project and covers only 2 CPUs, which AT&T distinguishes from DEC's other CPUs which are left under their prior license. So it's a special agreement, with another company entirely, not IBM, with terms of a completely different license from either the educational or the commercial ones we've seen before. It isn't at all comparable to IBM's license, so what it can possibly prove is a puzzlement. In particular, it doesn't have the "viral" "resulting materials" wording of Section 2.01 of the IBM license:

“Such right to use includes the right to modify such SOFTWARE PRODUCT and to prepare derivative works based on such SOFTWARE PRODUCT, provided the resulting materials are treated hereunder as part of the original SOFTWARE PRODUCT.”

This letter refers to clause 2.01, but note what it says about DEC's 2.01 clause:

"If in creating the derivative work you use the code in a SOFTWARE PRODUCT or use methods and concepts from a SOFTWARE PRODUCT or a SUBLICENSED PRODUCT, and such derivative work does not include any of our code or otherwise embody any of our methods and concepts, you may consider that you have all right, title and interest to such derivative work and such derivative work will not be subject to the provisions of the referenced Software Agreement."

Contrast this with SCO's position, stated baldly in the Memorandum, that all of AIX and Dynix are derivative works of Unix and that none of it can be donated to Linux:

"SCO does maintain that AIX and Dynix are subject to the restrictions of the licensing agreements because they are derivatives of UNIX, and consequently that IBM breached the license agreements by contributing to Linux any portion of those contractually-protected derivatives."

Evidently, SCO views AIX and Dynix in their entirety as being derived from UNIX, despite IBM, and Sequent, writing their own portions. Under this 1984 AT&T-DEC letter agreement, if it had anything to do with IBM, which it doesn't, IBM would own its own homegrown code and it would "not be subject to the provisions of the referenced Software Agreement". Isn't that what it says? Maybe SCO didn't notice that part. Probably it didn't read 12(d) either:

"DEC shall have the right to disclose any independently developed software which software is created without the use of or reference to AT&T Technologies proprietary and confidential information received under the referenced Software Agreement."

Isn't that pretty much what Mr. Wilson said? What's ours is ours and what's yours is yours? Even if SCO could find some discordant documents, where AT&T, as the judge in the BSDi case pointed out, didn't protect its code very well ("Version 32V source code has now been distributed, without notice, to literally thousands of licensees. Consequently, Plaintiff can have no valid copyright on 32V unless it can fit within one of the statutory or common law escape provisions. . . . Plaintiff cannot avail itself of any of these provisions."), despite periodic stabs in that direction, one of which resulted in the unfortunate (for USL) BSDi lawsuit, it isn't hard to figure out that you might find some small piece of apparently conflicting information if you go prospecting long enough and concentrate only on those periodic stabs.

I gather SCO is looking for any documents it can dredge up from those ineffective gestures in the direction of control, which might have worked if it weren't for the USL-Regents Settlement Agreement showing up and letting us know how it all worked out in the end.

The license refers to the SOFTWARE PRODUCT (whatever that is exactly -- some references are made to documents and attachments we don't have) and to AT&T modifying it to create a SUBLICENSED PRODUCT. Also, because this is the middle of negotiations over 2 CPUs for a particular project which is being distinguished from all other CPUs, which DEC apparently had a license to, some of the references are unknowable without further information. For example, what do you make of this clause:

12. (a) AT&T agrees that the same level of confidentiality that LICENSEE has agreed to for previous UNIX Operating System products shall be continued with the SOFTWARE PRODUCTS covered by the referenced Software Agreement.

See what I mean? Without knowing what DEC earlier agreed to, this is a meaningless clause today. Needless to say, it's an unusual arrangement between DEC and AT&T back in 1984, which has absolutely nothing to do with IBM and nothing to do with the Linux kernel, which wasn't even in existence in 1984. Everything, as Mr. Wilson has testified, changed in 1985 anyway.

Clause 12(b) makes reference to an attachment that specifies what was and what wasn't considered publicly known, but considering that we don't have the attachment and we're talking 1984, it seems irrelevant.

Why, then, did SCO scrape the bottom of the file cabinet and submit this document? It's what they had.

*******************************

[Letterhead]
AT&T Technology Systems
[Address, Phone]

O. L. Wilson
Manager, Software
Sales and Marketing

September 5, 1984

MR. LESTER S. GRODBERG
Digital Equipment Corporation
[Address]

Dear Mr. Grodberg:

Re: Proposed Software Agreement SOFT-00092 Between
Our Companies

This responds to your letters of August 15 and August 20, 1984 relating to this referenced Software Agreement. The numbered paragraphs correspond to the numbers in the attachment accompanying your letter of August 15. We understand that Digital Equipment Corporation (“DEC”) accepts AT&T’s current position on the issues raised by Mr. Grodberg’s letter of August 15, 1984 to achieve our mutual short term goal with the understanding that both parties will enter into further negotiations to resolve these issues.

1, and 2. We understand that you want the new agreement to cover only two CPUs that you have identified: accordingly, the attached Supplements 1 and 2 name only those two CPUs as DESIGNATED CPUs. It is agreed that licenses for UNIX* System V, Release 2.0 and UNIX Documenters Workbench** Software are granted for use on these two CPUs only for a fee of eight hundred dollars ($800.00). This reflects a four hundred dollar ($400.00) fee to upgrade both CPUs from UNIX System V, Release 1.1 to UNIX System V, Release 2.0 and UNIX Documenter's Workbench** Software, and distribution of the PDP 11/70 version, and a four hundred dollar ($400.00) additional distribution fee for the VAX 11/780 version. These fees are in lieu of the right-to-use fees listed on the Schedules attached to Supplements 1 and 2. It is understood and agreed that the terms and conditions of the referenced Software Agreement shall apply only to the two DESIGNATED CPUs and that the terms and conditions of previous agreements shall apply only to the two DESIGNATED CPUs and that the terms and conditions of previous agreements shall apply to the use of other CPUs for which DEC is licensed. With respect to the two DESIGNATED CPUs only, the referenced Software Agreement shall be regarded as replacing all applicable previous agreements. The SOFTWARE PRODUCTS licensed under the referenced Software Agreement may not be used on any CPUs

______________
*UNIX is trademark of AT&T Bell Laboratories.

**Documenter’s Workbench is a trademark of AT&T Technologies.

1

MR. LESTER S. GRODBERG

2.

except such two DESIGNATED CPUs and will be made available only to those DEC employees involved with the development of the UDA 50 driver. It is agreed that licenses to use UNIX System V, Release 2.0 and UNIX Documenter’s Workbench Software will be granted for other CPUs licensed under previous agreements for UNIX System V, Release 1.0 and 1.1 at no additional fee, provided that agreement can be reached on the terms and conditions of such licenses.

3. Regarding Section 2.01 of the referenced Software Agreement, you do not have all right, title and interest in derivative works based on a SOFTWARE PRODUCT if such derivative works include any of our code or embody any of the methods and concepts used in a SOFTWARE PRODUCT. Copies of such derivative works may be furnished to customers as SUBLICENSED PRODUCTS under a Sublicensing Agreement. If in creating the derivative work you use the code in a SOFTWARE PRODUCT or use methods and concepts from a SOFTWARE PRODUCT or a SUBLICENSED PRODUCT, and such derivative work does not include any of our code or otherwise embody any of our methods and concepts, you may consider that you have all right, title and interest to such derivative work and such derivative work will not be subject to the provisions of the referenced Software Agreement. If such derivative work does include any of our code or embody any of our methods and concepts you may have a property right in such derivative work to the extent of any modifications that you have added, but the exercise of that property right is subject to the terms of the Software and Sublicensing Agreement, including to the restrictions on the use of the SOFTWARE PRODUCT (for example, it must be kept in confidence).

4. It is not our practice to ship software to a licensee until an agreement has been fully executed and payment received by AT&T.

5. With Reference to Section 4.01, we agree that consent by AT&T to export SOFTWARE PRODUCTS for use on a DESIGNATED CPU of a SUBSIDIARY or other licensee of equivalent scope will not be unreasonably withheld.

6. AT&T maintains a listing of states which it believes apply sales and/or use taxes to software licensed therein. Any such tax is reflected on the invoice accompanying the software agreement. New York is not currently on that listing.

2

MR. LESTER S. GRODBERG

3.

7. We expect to keep archive copies of SOFTWARE PRODUCTS we license. Thus, if your rights under the referenced Software Agreement are ever terminated and you are subsequently required to gain access to a SOFTWARE PRODUCT licensed under such agreement, such access can be provided to our archive copy under appropriate protective arrangements.

We are willing to consider entry into an escrow agreement as you have described in your August 20 letter.

8. This point has been dropped.

9. AT&T’s grant of rights to modify SOFTWARE PRODUCTS under Section 2.01 does not relate to patents. Under Section 7.01 AT&T’s grant of patent licenses specifically excludes cases where the patent applies because of a modification of the SOFTWARE PRODUCT.

10. AT&T Technologies Software Sales and Marketing organization is not aware of any patent, trade secret or copyright infringement action against AT&T Technologies or its affiliates relating to the software covered by the referenced Software Agreement.

11. Regarding section 7.01, AT&T agrees that the permission referred to in this section will not be unreasonably withheld if such permission is requested in writing.

12. (a) AT&T agrees that the same level of confidentiality that LICENSEE has agreed to for previous UNIX Operating System products shall be continued with the SOFTWARE PRODUCTS covered by the referenced Software Agreement.

(b) All portions of the SOFTWARE PRODUCTS that have not been published shall be held in confidence. The documents listed in paragraph 2 (a) in the Schedule for UNIX System V, Release 2.0 except the System Release Description, the Portfolio and the International Release Description are considered to be published. None of the computer programs listed in paragraph 3 of such Schedule are considered to be published. The documents listed in Section 2(a) in the Schedule for UNIX Documenter’s Workbench Software, except the System Release Description, are considered to be published. None of the computer programs listed in paragraph 3 of such Schedule are considered to be published.

3

MR. LESTER S. GRODBERG

4.

(c) As needed, we will consent to disclosure to a consultant for limited purposes if the consultant agrees to the same obligations of confidentiality and other restrictions pertaining to the use of the SOFTWARE PRODUCT as those undertaken by you under the referenced Software Agreement.

(d) DEC shall have the right to disclose any independently developed software which software is created without the use of or reference to AT&T Technologies proprietary and confidential information received under the referenced Software Agreement.

(e) On further consideration we still are unwilling to set a time limit on confidentiality or agree to a limit on financial liability.

13. The verification process will be immediate in most cases and we expect it should never be longer than two business days.

14. (a) and (b) This matter is covered in the items 1 and 2.

(c) AT&T will give DEC ninety (90) days notice prior to any price changes.

If you agree with the above, please so indicate by signing and dating the attached copy of this letter and returning such copy to us.

Very truly yours,

AT&T TECHNOLOGIES, INC.

By ___[Signature] ___
For O. L. Wilson

Title Manager, Software Sales and Marketing

Date September 5, 1984

ACCEPTED AND AGREED TO:

DIGITAL EQUIPMENT CORPORATION

By __[Signature] __

Title Vice President, Mfg. & Eng.

Date September 18, 1984