UNITED STATES DISTRICT COURT DISTRICT OF NEW JERSEY UNIX SYSTEM LABORATORIES, INC., Plaintiff, (DRD) -against- BERKELEY SOFTWARE DESIGN, INC., and the Following lndividuals in their Collective Capacity as THE REGENTS OF THE UNIVERSITY OF CALIFORNIA: PETE WILSON, LEO T. MCCARTHY, WILLIE L. BROWN, JR., et al., Defendants. Civil Action No.92-1667 MEMORANDUM OF LAW ON BEHALF OF DEFENDANT BERKELEY SOFTWARE DESIGN, INC. IN FURTHER OPPOSITION TO PLAINTIFF'S MOTION FOR PRELIMINARY INJUNCTION JAMES H. FORTE (2248) SAIBER SCHLESINGER SATZ & GOLDSTEIN One Gateway Center Newark, New Jersey 07102-5311 (201) 622-3333 LESLIE A. FITHIAN HELLER, EHRMAN, WHITE & MCAULIFFE 525 University Avenue, 9th Floor Palo Alto, California 94301 (415) 326-7600 Attorneys for Defendant Berkeley Software Design, Inc. Scheduled Argument Date: January 22, 1993 TABLE OF CONTENTS TABLE OF AUTHORITIES . . PRELIMINARY STATEMENT . . . . . . . . . . . . . . . ... . . . . . . ARGUMENT . . . . _ . . . . . . . . . . _ . . . . . . . . . . . . . 2 I. USL HAS NO VALID COPYRIGHT IN THE 32V CODE . . . . . . . . . . . . . . . . . . . . . 2 II. THE HEADER FILES HAVE INDEED BEEN DISTRIBUTED IN BINARY RELEASES CONTAINING NO COPYRIGHT NOTICE . . . . . . . . . . . . . . . . . . . . 6 III. USL HAS NOT ESTABLISHED INFRINGEMENT, BY DIRECT EVIDENCE OR OTHERWISE . . . . . . . . . . . . . 8 A. USL Cannot Avoid The Substantial Similarity Test . . . . . . . . . . . . 8 B. USL Cannot Meet The Substantial Similarity Test . . . . . . . . . . . 8 C. The Similarities Upon Which USL Relies Are De Minimis . . . . . . . . . . . . . . . 16 IV. THE UNIVERSITY'S AGREEMENTS WITH USL PRECLUDE USL'S TRADE SECRET CLAIM . . . . . . . . . . . . 17 V. THE BALANCE OF HARM WEIGHS SHARPLY IN FAVOR OF DENIAL OF A PRELIMINARY INJUNCTION . . . . . . . . . . . . . . . . 19 CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 TABLE OF AUTHORITIES Cases Academy of Motion Picture Arts v. Creative House, 944 F.2d 1446 (9th Cir. 1991) . . . . . . . . . . . . . . . 5 Altai v. Computer Associates, 1992 WL 372273 (2d Cir. 1992) . . . . . . . . . . . . . . . . . . . . . . 8,13,1 Brown v. Tabb, 714 F.2d 1088 (11th Cir. 1983) . . . . . . . . . . . 5 DCI Computer Systems. Inc. v. Pardini, 1992 U.S. App. Lexis 29951; 1992 Copyright L. Rep. (CCH) . . . . . . . . . . . . . . . . . . Feist Publications. Inc. v. Rural Telephone Service Co., 499 U.S. ___, 113 L. Ed.2d 358, 111 S.Ct. 1282 (1991) . . . . . . . . . . . 8 Hubco Data Products Corp. v. Management Assistance. Inc., 219 U.S.P.Q. 450 (D. Idaho 1983) . . . . _ . . . . . . . . . . . . . . . . . . . . 2 S.A.S. Institute, Inc. v. S&H Computer System. Inc., 605 F. Supp. 816 (M.D. Tenn. 1985) . . . . . . . . . lS SEGA Enterprises v. Accolade, 977 F.2d 1510 (9th Cir. 1992) . . . . . . . . . . . . . . . 9,13,17 Shapiro & Son Bedspread Corp. v. Royal Mills, 568 F. Supp. 72 (S.D.N.Y. 1983) . . . . . . . . . . . . . . . . . 7 Videotronics, Inc. v. Bond Electronics, 586 F. Supp. 478 (D. Nev. 1984) . . . . . . . . . . . . . . . . . . 7 White v. Kimmel, 193 F.2d 744 (9th Cir.), cert. denied, 343 U.S. 951 (1952) . . . . . . . . . . . . . . . . . 2 Other Authorities Compendium of Copyright Office Practices, Section 325.02(c) 3 Nimmer on Copyright Section13.03(A) _ . . . . . . . . . . . . . . . 3 Nimmer on Copyright, Section13.01(B) . . . . . . . . . . . _ _ . . 3 3 Nimmer on Copyright, Section13.03(F) _ _ . . . . . . . . . . . . . PRELIMINARY STATEMENT USL has submitted over eighty pages in reply to BSDI 's opposition, much of which consists of new matter. USL appears to have abandoned most of its trade secret claim, now arguing only that the actual disclosure of any 32V code to nonlicensees violates the University's license agreements with USL. USL overlooks the provision in those agreements that explicitly excludes from the University's obligation of confidentiality any information that is generally known. Moreover, USL has failed to address the evidence submitted by BSDI and the University that all of the features USL has relied on are generally known. The new items discussed in the Carson affidavit likewise are unable to withstand scrutiny. In seeking to salvage its nonexistent copyright claim, USL attempts to persuade this Court to abandon the well-established test for limited publication, even though courts have invariably upheld that test as applicable, both before and after the 1976 Copyright Act. All of the cases USL cites acknowledge this as the proper test. Nor can USL -- a company set up for the specific purpose of marketing UNIX source code for financial gain -- legitimately claim that it has offered its source code only to a limited group for a limited purpose. Since USL has no copyright in 32V, the Court need not address the other issues with respect to USL's claim of infringement. Even if there were a copyright, moreover, USL has not established infringement. Contrary to USL's assertions, the deposition testimony of Kirk McKusick provides no admission of infringement. Even when there is direct evidence of copying, the substantial similarity test must still be met. USL has made no showing of quantitatively or qualitatively substantial similarity between 32V and Net2 or BSD/386. Even if USL could establish substantial similarity, the scope of the injunction it is seeking is unconscionably overbroad. USL has no response to this charge other than to say that it is "premature." USL's Reply Brief at 2, n.3. In fact, it is USL's request for an injunction that is premature, since it has not met any of the requirements for obtaining such relief. ARGUMENT I. USL HAS NO VALID COPYRIGHT IN THE 32V CODE Ignoring all the case law to the contrary, USL argues that this Court should ignore the well-established test for compliance with the limited publication doctrine. USL has failed to cite a single case to support its argument that the Court should ignore this test. Indeed, even in the cases USL relies on, the courts invariably have adopted the three-pronged test set forth in White v. Kimmel, 193 F.2d 744 (9th cir.), cert. denied, 343 U.S. 957 (1952), in which the Court held that restrictions on redisribution alone are not sufficient to meet the limited publication doctrine. See, e.g., Hubco Data Products Corp. v. Management Assistance, Inc., 219 U.S.P.Q. 450, 455 (D. Idaho 1983) ("[The White v. Kimmel definition] of 'limited publication' remains valid for judging conduct occurring both before and after the copyright Act became effective in 1978"). [fn1] [fn1] If licensing restrictions were sufficient to make a publication limited, as USL argues, software would never be "published," since it is standard practice to license software pursuant to various restrictions on its use and further distribution. Yet, unlike the 32V source code, most software is distributed with a copyright notice. USL would have this court ignore two of the three requirements for a limited publication -- the limited group and limited purpose requirements -- precisely because USL realizes that it cannot possibly meet those requirements. USL's 32V code has not been offered to a limited group for a limited purpose, but in its own words, has been marketed "to the commercial world". (see AT&T Technical Journal, p. 1573, Forte Decl. Exh. P; Thiel Dep. T. 31-4 to 37-5, Forte Decl. Ex. A). USL has struggled to find examples of instances in which it turned down a prospective customer, arguing that it had the discretion to enter into agreements or not. When a company markets a product, it always has the option of turning away a particular customer. But the fact that a few customers have been turned down for business reasons is hardly sufficient to make it a limited publication. To hold otherwise would be to turn the doctrine of limited publication on its head: Rather than making its source code available to only a limited group, USL *has withheld it* only from a limited group, USL's own documents belie its current attempt to portray its source code offering as limied. For example, USL's [TEXT OMITTED HERE] .... Likewise, as noted above, in a journal published by AT&T, AT&T represented that UNIX source code has been "available to the commercial world." (See AT&T Technical Journal, p. 1573, Forte Decl. Exh. P). There was no hint in these documents that only a narrowly selected group within the commercial world could obtain the source code, and [TEXT OMITTED] .... Thus, regardless of whether USL (like anyone selling a product) had the discretion to refuse prospective customer, the reality is that it has rarely done so. USL also misconstrues the limited purpose requirement, focusing only on the purposes for which its licensees may use the code. Regardless of those restrictions, since USL's purpose in distributing the code has been to commercially exploit the code, it cannot meet the limited purpose requirement. See, e g., DCI Computer Systems. Inc. v Pardini, 1992 U.S. App. Lexis 29951; 1992 Copyright L. Rep. (CCH), par. 27,005 (9th Cir.) (limited purpose test not met where purpose of distribution was financial gain); Academy of Motion Picture Arts v Creative House, 944 F.2d 1446, 1453 (9th Cir. 1991) (limited purpose requirement met because commercial gain was incidental rather than direct); Brown v.Tabb, 714 F.2d 1088, 1091- 92 (11th Cir. 193) (commercially exploited work did not meet limited purpose test). USL correctly points out that the purpose of the copyright law is to encourage authors to disseminate ideas. The point of the limited publication doctrine is to distinguish between instances in which the author of a work has chosen to take advantage of the benefit of commercial exploitation of copyrighted works and instances in which the author has instead chosen to forego such benefits by not seeking financial gain from distribution of work. Where an author chooses to widely disseminate his work for financial gain, as USL as done, he must meet the statutory requirements in order to claim copyright. USL, however, wants to have its cake and eat it too. One cannot market a product to the commercial world and simultaneously claim it is an unpublished work, simply to avoid the limitations and requirements imposed by copyright law. Thus, USL's argument on the limited purpose prong misses the point entirely. II. THE HEADER FILES HAVE INDEED BEEN DISTRIBUTED IN BINARY RELEASES CONTAINING NO COPYRIGHT NOTICE USL provides examples of instances in which binary distributions did have copyright notices, and states that it is unaware of distributions without a copyright notice. In fact, USL's binary distribution agreements from the early 1980s [TEXT OMITTED] USL has submitted no evidence contradicting the testimony of Keith Bostic and Kirk McKusick that no copyright notices appeared in the early binary releases they saw. (See Bostic and McKusick Joint Decl., 28.4). In any event, USL relied solely on copyright notices that appeared on the "media" in which the code was contained, ignoring the fact that ordinary users of such software would never have seen the media, but rather, were permitted to access the code directly on the computer through timesharing arrangements, i.e., the software was to be used on computers which could be accessed by many different users who would never see the "medium" on which the code was originally stored. [fn2] All citations to the Second Supplemental Declaration of James H. Forte shall be referred to as "Forte Sec. Supp. Decl." To be effective, notice had to be given "in such a manner and location as to give reasonable notice" to the ordinary user. 17 U.S.C. Section 401 (former version). It is not sufficient to put a notice on a box that those actually using the software would never see. See Shapiro & son Bedspread Corp. v. Royal Mills, 568 F. Supp. 972, 975 (S.D.N.Y. 1983) (notice on a wrapper or container which might not be seen by user is inadequate notice); Videotronics, Inc. v. Bond Electronics, 586 F. Supp. 478, 42 (D. Nev. 1984) (notice appearing randomly on videoscreen is inadequate because it fails to give notice to each user). Since even USL does not dispute the fact that these header files have been published and has failed to refute the evidence that they contained no visible copyright notice, USL cannot claim copyright protection in such files. In any event, these same header files also were included in the source code distributions of 32V. It is undisputed that such distributions contained no copyright notice. With respect to such distributions, USL relies only on its contention that its distribution of 32V source code was merely a limited publication -- an assertion that finds no support in the facts or the case law, as explained above. III. USL HAS NOT ESTABLISHED INFRINGEMENT, BY DIRECT EVIDENCE OR OTHERWISE Even assuming the existence of a valid copyright, USL has not shown a likelihood of success on its copyright claim. A. USL Cannot Avoid The Substantial Similarity Test USL erroneously argues that evidence of direct copying renders unnecessary the "substantial similarity" analysis. his argument was rejected recently by the United States Supreme Court in Feist Publications, Inc., v. Rural Telephone Service Co., 113 L. Ed.2d 358 (1991). As the Supreme Court held in that case, "[n]ot all copying ... is copyright infringement." See Altai v. Computer Associates, 1992 WL 37223, at 20 (2d Cir. 1992) ("once some amount of copying has been established, it remains solely for the trier of fact to determine whether the copying as 'illicit'; "We have subsequently described this method of inquiry as 'merely an alternative way of formulating the issue of substantial similarity"). As Nimmer explains, "substantial similarity between the plaintiff 's and defendant's works is an essential element of actionable copying." 3 Nimmer on Copyright Section 13.03[A]. Thus, even if "one can conclude that, as a factual proposition, copying took place, ... the question still remains whether such copying is actionable." Id. 13.01[B]. Thus, USL may not avoid the need to show substantial similarity in protectible expression. B. USL cannot Meet the Substantial Similarity Test The testimony of Dr. McKusick, which USL quotes out of context, provides no evidence of infringement. As explained by Dr. McKusick in his Second Declaration,[fn3] his testimony regarding reliance on the existence of pseudocode in reviewing Net2 involved a determination as to whether code implementing algorithms used in 32V could be included, not whether 32V code itself could be included. (See Second Bostic and McKusick Decl., Para 2.3). Algorithms are not even copyrightable (See Compendium of Copyright Office Practices, 325.02(c), Forte Decl. Exh. W); see also 3 Nimmer on Copyright, Section 13.3[F], at 13-85, n.289. To the extent any code was included based on the existence of pseudocode, it was not the 32V code, but code written by the University and its contributors. (See Second Bostic and McKusick Decl., Para. 2.3). USL also cites testimony of Dr. McKusick regarding his familiarity with 32V code. But mere familiarity with a work does not make one's later works infringing. Indeed actual reference to or use of a work in developing one's own code does not amount to copyright infringement. Such uses are exactly what the copyright law is designed to foster. See, e.g., SEGA Enterprises v. Accolade, 977 F.2d 1510, 1527 (9th Cir. 1992). Thus, this testimony is irrelevant to the question of whether NET2 or BSD/386 would infringe a copyright in 32V (if one existed). McKusick also testified that in programming one might reuse a four- line piece of code, much like one might reuse a footnote. But McKusick was testifying hypothetically about common practices in programming, and not testifying to a particular event. (See Second Bostic and McKusick Decl., 2.4); see also 3 Nimmer on Copyright, Section 13.03[F], at 13-96 (Most programmers rely on a number of traditional solutions to recurring problems in their programming"). Thus, USL's reliance on this testimony as evidence of copying is misplaced. [fn3] All citations to the Second Declaration of Keith Bostic and Marshall Kirk McKusick, which is being submitted on behalf o the defendant Regents of the University of California, shall be referred to as "Second Bostic and McKusick Decl." In his Second Declaration, McKusick does note that after the Release of Net2, he discovered that a total of 56 lines of 32V code had been inadvertently included in the Net2 Kernel, scattered in 5 files; but he also demonstrates that those lines are not significant either quantitatively or qualitatively and they thus do not amount to copyright infringement, nor do they provide any basis for prohibiting use of other code, as to which there has been no showing of copying, de inimis or otherwise. Since the Net2 distribution was a combination of many programs and subprograms written by many people over a period of many years, even if USL could show copying of substantial expression in a particular piece of code, this would have no bearing on whether the remaining code -- written by different people at different times - was copied. None of the testimony of Dr. McKusick upon which USL now relies concerned the particular features relied on by USL in it motion and addressed in David Kashtan's Affidavit. Thus, USL's attempt to dismiss the Kashtan Affidavit in light of this testimony is unavailing. While USL states that Kashtan concluded that "all of the remarkable textual similarities" found by Dr. Carson were standard and well-documented, in fact, Kashtan found those similarities unremarkable, and unreliable as proof of copying, given the fact that they are well-documented and widely known and used. Out of the more than 1.5 million lines of code in BSD/86, in his initial affidavit, Dr. Carson pointed to no instances of similarity in the working code itself, other than a few isolated lines in two functions within a single file, which, as noted in the Kashtan Affidavit and in the Joint Declaration of Bostic and McKusick, evidence only the use of standard programming techniques to implement functions described in the available reference materials. There is no evidence to support Dr. Carson's conclusion that such code was copied. Since defendants have rebutted all of the items relied on by Dr. Carson in his initial affidavit, he now attempts to rely on new evidence which has never before been disclosed to BSDI. Carson contends that he has provided only examples, but other than those examples he has proffered nothing more than unsupported conclusory statements. Nor is there any reason to believe that any such unidentified similarities would be any more probative than those USL has chosen to identify. To the extent USL has evidence it wants considered in this motion, it must place that evidence before the Court, and provide BSDI a fair opportunity to respond to the assumptions on which Carson's conclusions are based. Such conclusory allegations of unidentified similarities are not sufficient to support a motion for preliminary injunction. In any event, USL's new evidence appears no more reliable than what it presented the first time around: ACTUAL CODE. In actual working code, USL has identified only [TEXT OMITTED] identical lines in a particular file, [TEXT OMITTED] with no analysis as to whether those lines are significant in any way. (See Second Bostic and McKusick Decl., Para 6.1). In reality, those lines provide no particular value and are easily replaceable. Although they were inadvertently included in the Net2 release and can easily be removed from the release, they are a classic example of de minimis copying and thus provide no basis for a finding of infringement. In an attempt to avoid disclosing how little similarity actually exists in the code, USL's expert makes vague references to other unidentified similarities. In reality, only 56 lines out of the 230,995 lines in the Net2 kernel re common to 32V. (See Second Bostic and McKusick Decl., 2.1-.1.5). These common lines do no appear together; but rather are interspersed in only 5 of the Kernel files. In each of these instances, the small piece of common code is completely trivial and cannot be considered a taking of any substantial protectible expression from 32V. (Id.). COMMENTS. USL also relies on a small number of identical comments, which consist of short, fragmentary phrases describing the function performed by a piece of code. They are not part of the code itself, have no role in the operation of the program and are not protectible as "original expressions of authorship. " (See second Bostic and McKusick Decl. Paras. 6.2, 14.1.A, 15.5; see also BSDI's Memorandum of Law in opposition to USL's motion and cases cited therein). Moreover, the range of expression in such comments is extremely limited since they are entirely descriptive of a narrow idea, i.e., the function performed by a small piece of code. For this reason as well, they are not protectible. See Altai, 1992 WL 372273 at 14; SEGA Enterprises, 977 F.2d at 1524, n 7. CPIO. USL also relies on use of the CPIO file in Net2 as evidence of copying. Curiously, although Carson characterizes it as the most "extreme" example of copying, he failed to even mention it in his initial Affidavit. He failed to mention it for good reason, since it was included in Net2 only after USL representatives urged the University to use CPIO in its releases in place of another routine called TAR, which the University had previously used to perform the same function. USL made this request specifically because it wanted CPIO to become a standard . The University had no particular interest in changing from the AR format to the CPIO format, since the TAR format is superior. At the request of USL, however, the University agreed to use CPIO instead. (See Second Bostic and McKusick Decl ., 16 3 ) . These facts were well-known to USL and its counsel since Kirk McKusick testified to them at his deposition, yet USL never mentions them in its reply papers, disingenuously implying that the presence of CPIO is simply an example of surreptitious copying. Thus, USL's reliance on the inclusion of CPIO as evidence o copying is a deliberate misrepresentation of the facts. HEADER FILES. With the exception of a couple of comments and the few lines of identical code discussed above, all the features USL relies on are in header files. For the most part, USL relies on nothing more than common mnemonics that are well-known and well- documented. These interfaces are provided in header files and in a variety of programmer's reference materials, precisely so that they can be used in developing compatible software, and they have historically been so used with no objection from USL or its predecessors. (See Second Bostic and McKusick Decl., 6.4; Bostic and McKusick Decl., 28.3 28.7). Thus, such use cannot form the basis for an infringement action. Although Richard Hamilton admits that [TEXT OMITTED] In making this argument, he ignores the fact that those programs are more limited in scope than Net2, do not contain all o the same functionality, and make no attempt to be compatible with device drivers written for use with UNIX-compatible programs. (See Second Bostic and McKusick Decl., Paras.17-34) He ignores, however, numerous other documents that identify the remaining.file names. The other arguments made hy Carson and Hamilton with respect to the header files are equally specious as explained in detail in the Supplemental Declaration of Bostic and McKusick. (See generally Second Bostic and McKusick Decl.) C. The Similarities Upon Which USL Relies Are De Minimis USL makes a half-hearted attempt to argue that the specifics relied on by Dr. Carson are not de minimis, but neglects to provide any facts to support this conclusory assertion. In reality, USL has identified only a small number of lines of working code that are identical or similar to lines in 32V. As explained in BSDI's opposition papers, some of these lines of similarity are to be expected in light of standard programming techniques and other available reference materials, and provide n evidence that those lines were copied, much less any evidence f copying elsewhere in Net2. The others, as explained in the Second Bostic and McKusick Declaration, were inadvertently included and are trivial both quantitatively and qualitatively. USL's reliance on S.A.S. Institute, Inc. v. SH Computer System. Inc_, 605 F_ Supp. 816 (M.D. Tenn_ 1989), is misplaced. In that case, the court found that those particular lines were qualitatively significant, not quantitativelysignificant. See S.A.S. Institute, 605 . Supp. at 829. Likewise, in the other cases on which USL seeks to rely, there was a showing that the specific code copied was qualitatively significant. Although USL argues that the kernel as ,a whole is significant, it makes no attempt to argue that the specific lines of identical or similar code at issue in this case are significant. In reality they have no special significance and are easily replaced with no effect whatsoever on the operation of the files at issue. (See generally Second Bostic and McKusick Decl.) Lacking any similarity in the actual code that is not completely trivial, Carson relies instead on similarities in unprotectible function names and constant definitions, which as explained in BSDI's and the University's opposition papers and supporting affidavits, consist of well-known abbreviations or "mnemonics," commonly used in UNIX-compatible programs. Dr. Carson provides no support for an argument that such names are either a quantitatively or qualitatively significant portion of he 32V source code. Thus, there is no basis for a finding of infringement based on the use of such mnemonics. See Altai, supra, 1992 WL 372273 at 22 (use of a few similar parameter lists and macros "did not warrant a finding of infringement given their relative contribution to the overall program"); see also SEGA Enterprises, 977 F.2d a 1516, 1524, n.7 (admitted copying of header file containing interface needed for compatibility did no give rise to a claim for copyright infringement). IV. THE UNIVERSITY'S AGREEMENTS WITH USL PRECLUDE USL'S TRADE SECRET CLAIM At least for purposes of this motion, USL has not disputed the defendants' interpretation of its license agreements with USL, and appears to have abandoned its claim that any familiarity with the UNIX source code is sufficient to taint programmers writing their own code. [fn4] Thus, the only issue is whether any trade secrets contained in USL's source code were disclosed to non-licensees. Faced with overwhelming evidence to the contrary, USL also appears to have abandoned the contention that the methods and concepts used in UNIX source code are trade secrets . [fn4] Although USL has modified its prayer for relief in its proposed Second Amended Complaint, in its proposed order submitted with this motion, USL requests the Court to enjoin BSDI from employing any programmers that have had access to UNIX source code. USL's only remaining argument -- that disclosure of *any* USL code to non-licensees violates the University's license agreement -- totally ignores Section 7.05(a) of the E-SOFT-0089 Agreement (and the identical provisions in the earlier agreements). (See Exh. D-25 to the Bond Deposition Transcript, Forte Decl. Exh. F). That section expressly excludes from the duty of nondisclosure any licensed information that is or later becomes generally known. Thus, the University had no obligation to limit disclosure to generally known code, methods or concepts. As established in the Bostic and McKusick Declarations and the Kashtan Affidavit, USL has not demonstrated that any of the similarities relied on by Dr. Carson were secret at the time Net2 was released. Rather, all of those elements were well-documented in publicly available reference materials. USL's argument that defendants have not shown that those materials were actually relied on in developing Net2 misses the point, since the University's obligation of nondisclosure did not apply to any 3-14; Bostic and McKusick Decl., Paras. 2- 2.13; Second Bostic and McKusick Decl., Para. 15.4). USL's arguments regarding the advantages one would have had in working on such code as a result of access to or familiarity with USL licensed code are simply irrelevant in that the license agreements prohibit only disclosure of works that actually contained USL proprietary information, and exclude from the obligation of confidentiality any information that is or becomes generally known. Since USL chose to disclose its source code to student programmers throughout the world, encouraged its use among such students for educational purposes, and has licensed its source code to all the major competitors in the computer industry, as a practical matter, mere familiarity with UNIX source code provides no special advantage not shared by the industry as a whole. V. THE BALANCE OF HARM WEIGHS SHARPLY IN FAVOR OF DENIAL. OF A PRELIMINARY INJUNCTION USL admits that only a small portion of the Net2 release is even accused in this motion. Even within the kernel, USL points only to a few lines of actual code in a couple of files, which lines are both quantitatively and qualitatively insignificant. USL's concern about comments is ludicrous, since the use of any particular phrase to describe the function performed by a portion of the code provides no advantage whatsoever. USL's real goal is to bar any significant competition by preventing the use of standard file names and mnemonics in interfaces that are required in order to make the operating system compatible with other computer programs -- mnemonics that are not even protectible under copyright law and are well-documented and available for use by all programmers. Moreover, since USL has no copyright in the 32V code, the court need not even address these issues. Based on such flimsy arguments, USL asks this Court to put BSDI out of business (which would mean instant unemployment for all of its employees), without even granting it the right of a trial USL asks the Court to do this simply to sen a message to the thousands of individuals that already have the Net2 release. No such message is necessary. The existence of USL's challenge to Net2 is already well- known in the industry, and has had an inhibiting effect. USL will suffer no harm if BSDI is permitted to remain in business long enough to complete discovery and present its case at trial. On the other hand, a grant of USL's motion will not only destroy BSDI and prevent it from having the financial means to defend itself, but also will cast a pall over the entire industry by locking up information that has long been in the public domain. Thus, USL cannot show any compelling need for preliminary relief, and particularly cannot show any harm outweighing the harm to BSDI should the far-reaching injunction it seeks be granted. CONCLUSION In seeking to enjoin BSD/386, USL attempts to prevent BSDI from using thousands of files as to which USL has submitted no evidence whatsoever, most of which it admits are not even accused. Even within the kernel, USL accuses only a small number of files, with no showing that any substantial, protectible expression has been copied. Since USL does not even have a valid copyright in 32V, moreover, these issues are moot. USL's trade secret claim likewise is completely unsupported. USL's only remaining argument ignores the plain language of the licensing agreements, which do not require that the University limit disclosure of information that is already generally known. For all the reasons set forth above and in the the papers submitted by BSDI and the University in opposition to USL's motion, USL's request for a preliminary injunction should be denied. Respectfully submitted, SAIBER SCHLESINGER SATZ & GOLDSTEIN Attorneys for Defendant Berkeley Software Design, Inc. By: JAMES H. FORTE A Member of the Firm Dated: January 20, 1993 Of Counsel: Leslie A. Fithian, Esq. Heller, Ehrman, White & McAulife 525 University Avenue, 9th Floor Palo Alto, California 94301