Appellant Plains Cotton Cooperative Association of Lubbock, Texas, appeals the denial of its application for a preliminary injunction against appellees Goodpasture Computer Service, Inc., Peter H. Cushman, Richard R. Fisher, William James Godlove, and Clarence Michael Smith, to prevent them from marketing, distributing, or otherwise using software allegedly stolen or copied from appellant. Because we agree with the district court that appellant has not satisfied its burden of demonstrating irreparable harm or substantial likelihood of success on the merits, we affirm the denial of the application for a preliminary injunction.

Plains Cotton Cooperative Association of Lubbock, Texas ("Plains") is a non-profit agricultural cooperative comprised of approximately twenty thousand cotton farmers residing in Texas and Oklahoma. Its purpose is to assist its members in the growing and marketing of cotton. Toward that end, Plains developed a computer software system, Telcot, designed to provide its members with information regarding cotton prices and availability, with accounting services, and with the capability to consummate actual sales electronically.

The Telcot "system" comprises functional specifications outlining what the software can accomplish, design specifications determining how the software accomplishes those functions, programs implementing the design specifications, and documentation detailing the programming and specifications so that future programmers can understand and modify the software. The programs, written in human-readable computer language, are called "source code;" translated into a computer-readable language, they are called "object code." The Telcot system is used by cotton farmers, ginners, and buyers through terminals connected to Plains' large central computer by telephone lines. These terminals allow the system's users to retrieve the desired cotton market information from Plains, but they provide no access to the documentation, programming, design or functional specifications of the Telcot system.

Telcot was first marketed in 1975, and improvements and refinements in the software have been made continuously since that date. The system was developed at Plains by a team of programmers that included appellees Cushman, Fisher, Godlove, and Smith ("employee/appellees"). None of these employees was required to sign confidentiality agreements as a condition of his employment with Plains.

In 1979, Plains' general manager Dan Davis left the company in order to start a new venture called Commodity Exchange Service Company ("CXS"). Davis had been heavily involved with the creation of Telcot, and he wanted to expand its capabilities by adapting Telcot for use on personal computers. CXS and Plains entered into an agreement, on July 5, 1979, for the development, refinement, and marketing of a personal-computer version of Telcot. Under the agreement, any modifications, developments or enhancements of Telcot would be owned jointly by Plains and CXS.

On April 13, 1984, CXS hired Cushman, Fisher, Godlove and Smith away from Plains, and put all four of them to work on the Telcot personal-computer project. When Godlove left Plains, he copied and took possession of a complete tape record of the Telcot source code detailing Telcot's programming. Five days later, Plains gave CXS written notice that it intended to terminate their contract. Subsequently, the parties agreed that their contract would terminate on April 18, 1985, and that Plains would have an irrevocable option through April 18, 1986 to purchase CXS's interest in various programs. Plains never exercised that option.

After working at CXS for approximately one year, the four employees/appellees succeeded in designing a personal-computer version of Telcot, and they produced a document setting out design specifications for the new software. The preliminary nature of the design and the comprehensiveness of the design document are disputed by the parties in this appeal. No programming was actually written for the new software.

In March, 1985, CXS filed for bankruptcy, and the four employee/appellees began to search for other employment opportunities. At that same time, appellee Goodpasture Computer Service Inc. ("Goodpasture"), a computer service company located in Brownfield, Texas, began discussions with CXS and its employees concerning various possible joint ventures or working relationships. Goodpasture knew that former Plains employees were working at CXS to create a personal-computer version of Telcot. No agreement was made between Goodpasture and CXS, but Goodpasture did succeed in hiring away from CXS appellees Cushman, Fisher, Godlove, and Smith.

The four employee/appellees signed employment agreements with Goodpasture in which they agreed not to breach any confidences of their former employers, Plains and CXS, while working for Goodpasture. In violation of this employment agreement, however, Fisher brought to his new job a computer diskette containing Telcot programming designs. Twenty days after arriving at Goodpasture, the four former Plains employees had completed a design of a personal-computer version of a cotton exchange program, designated "GEMS" by Goodpasture. By November, 1985 Goodpasture began to market GEMS in a rough, incomplete form. At the time of the preliminary injunction hearing, the software was still not fully operational.

GEMS is very similar to Telcot on the functional specification, programming, and documentation levels. In fact, several pages of the GEMS design manual appear to be direct copies of pages from the design manual appellees created at CXS. The main difference between the two systems is that Telcot is designed to work on a mainframe computer, whereas GEMS is designed for a personal-computer. Appellees allege that, with one exception, they did not copy programs used for Telcot, but instead "drew on their knowledge of the cotton industry and expertise in computer programming and design gained over a number of years." The exception is appellee Fisher's admission that he did copy one Telcot subroutine in programming GEMS. When Goodpasture discovered the copying on February 7, 1986, the subroutine was replaced, and Fisher was discharged.

On January 15, 1986, Plains filed suit against Goodpasture and its employees Cushman, Fisher, Godlove, and Smith, seeking damages and injunctive relief for misappropriation of trade secrets, unfair competition, conversion, breach of confidential relationship, and trade dress infringement. After registering its copyright in the Telcot computer programs and associated manuals, Plains amended its complaint to include claims for copyright infringement.

Plains made a motion for a preliminary injunction based on its copyright and trade secret claims, and a hearing was held on February 7 and 10, 1986. Both sides presented expert testimony on the issue of whether GEMS was copied from Telcot. On February 17, the district court denied the motion, and Plains subsequently instituted this appeal.

In order to secure a preliminary injunction, the movant has the burden of proving four elements: (1) a substantial likelihood of success on the merits; (2) a substantial threat of irreparable injury if the injunction is not issued; (3) that the threatened injury to the movant outweights any damage the injunction might cause to the opponent; and (4) that the injunction will not disserve the public interest. Enterprise International, Inc. v. Corporacion Estatal Petrolera Ecuatoriana, 762 F.2d 464, 471 (5th Cir.1985). These four elements are mixed questions of fact and law. Apple Barrel Productions, Inc. v. Beard, 730 F.2d 384, 386 (5th Cir.1984). In reviewing the findings of fact made by the district court, we will uphold those findings unless they are clearly erroneous. Id. In contrast, the court's conclusions of law "are subject to broad review and will be reversed if incorrect." Commonwealth Life Insurance Co. v. Neal, 669 F.2d 300, 304 (5th Cir.1982).

After reviewing findings of fact and conclusions of law, however, we note that ultimately a "district court's decision to grant or deny a preliminary injunction lies within its discretion." Enterprise International, 762 F.2d at 472; Apple Barrel Productions, 730 F.2d at 386. Thus, we are bound to affirm the judgment of the district court absent an abuse of discretion.

In denying the motion for a preliminary injunction, the district court concluded that appellant had failed to satisfy two of the four prerequisites for injunctive relief: a showing of irreparable harm and a showing of a substantial likelihood of success on the merits. Appellant contends that the district court abused its discretion (1) by using the wrong legal standard in analyzing the copyright infringement claim, and (2) by not addressing the misappropriation of trade secrets claim.III.

Appellant argues that the district court applied the wrong legal standard in determining that appellant had not demonstrated a sufficient likelihood of success on the merits of its copyright infringement claim. Copyright infringement is shown by proof that the injured party owned copyrighted material and that the infringer copied that material. Miller v. Universal City Studios, 650 F.2d 1365, 1375 (5th Cir.1981). Copyright ownership, in turn, is shown by proof of originality, copyrightability, and compliance with applicable statutory formalities. Apple Barrel Productions, 730 F.2d at 387. Since direct evidence of copying is often difficult to find, copying can be shown by proof of both access to the copyrighted material and substantial similarity between the two works. Miller, 650 F.2d at 1375; M. Nimmer, 3 Nimmer on Copyright Sec. 12.11[D] (1985).

Appellant asserts that the district court failed to consider evidence of substantial similarity between GEMS and Telcot. This argument is based on the district court's apparent reliance on the testimony of two expert witnesses, which testimony, according to appellant, conflicts only on the issue of direct copying.1 But direct copying is not the central issue in the case, appellant argues; it is "organizational copying": the reproduction of the organizational structure of a software system, outlined generally in the software's design specifications. Appellant asserts that because the expert testimony concerning organizational copying does not conflict, and because the court decided the preliminary injunction motion based on the conflict in testimony of the experts, then the court must have made its decision solely on the basis of whether direct copying occurred. The decision, therefore, ignored the allegedly uncontroverted evidence of substantial similarity resulting from organizational copying.

We do not accept the contention that the district court ignored the evidence of organizational copying and focused exclusively on evidence of direct copying. While the opinion by the district court is admittedly lean, it is open to interpretations other than the one profferred by appellant. The district court specifically based its finding "upon the hearing and the part's [sic] post-hearing briefs," not just on the testimony of the experts. A complete review of the record reveals testimony by several witnesses for appellees that the similarity between the two programs exists on a level not protected by appellant's copyright. Appellees' expert witness, James Bruce Walker, testified that appellees did not copy Telcot. In his opinion, appellant's programs were too large to have been copied and modified in the amount of time appellees took to create GEMS. Walker testified that appellees must have created the software based on their personal skills as computer programmers: "I believe they can take their knowledge of the cotton industry and repeatedly recreate a similar vehicle."

Appellant argues that Mr. Walker was not competent to testify as to whether "copying" occurred because he did not look for organizational copying. The record reveals, however, that Mr. Walker simply was not familiar with the term "organizational" to describe copying on a level broader than verbatim, line-by-line copying. However, Mr. Walker did testify on direct examination that he compared the "overall structure of these programs," and on cross-examination that he "expanded his study" of the programs from the narrow search for verbatim copying. The district court was entitled to rely on both of these statements in interpreting Mr. Walker's expert opinion that GEMS was not "copied" from Telcot.

Further support can be found in appellee William James Godlove's testimony that he did not rely on any material belonging to Plains or CXS during his employment at Goodpasture, that a mainframe program such as Telcot could be altered to run on a personal computer only with enormous changes so that rewriting the programs would be faster than modifying them, and that the residual familiarity with Telcot on which he relied when designing GEMS was "just experience and industry knowledge." Finally, witness Stinson Stokes Smith testified that the subroutines he wrote for Goodpasture were written without reference to any material relating to any of Plains' or CXS's software systems. This testimony is sufficient to support a tentative finding that appellees did not copy appellant's software. The district court's conclusion that appellant did not demonstrate a sufficient likelihood of success on the merits is not clearly erroneous.

Nor is the finding that appellant failed to demonstrate irreparable harm clearly erroneous. At oral argument, appellant relied on Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240 (3rd Cir.1983), cert. denied, 464 U.S. 1033, 104 S. Ct. 690, 79 L. Ed. 2d 158 (1984), to assert that irreparable injury is presumed upon a finding of likelihood of success on the merits. That rule, however, is not established in this circuit.2 On the contrary, we have made it clear in our decisions that preliminary injunctions will be denied based on a failure to prove separately each of the four elements of the four prong test for obtaining the injunction. See, e.g., Apple Barrel Productions, 730 F.2d at 389; Southern Monorail Co. v. Robbins & Myers, Inc., 666 F.2d 185, 186 (5th Cir.1982).

Appellant did not persuade the district court nor does it persuade us that any harm stemming from the alleged infringement of its copyright is not compensable in damages. In its brief and at oral argument, appellant limited itself to the assertion of a presumption of irreparable injury. The facts supporting appellant's position, however, seem as elusive as the legal principle upon which appellant relies. Appellant's copyright covers a piece of commercial software, and appellees are engaged in a for-profit enterprise of marketing that alleged infringing program. On the surface, this situation does not seem to present many possibilities for the infliction of unquantifiable damages. Finally, appellee points out that appellant is entitled to assess its customers a "sold around" charge for cotton not sold through its system. Appellant's vice president testified that, "if a cooperative gin had a Telcot agreement with us and the bale did not sell through the Telcot system, that's fine, but we get the dollar and a half and it is called sell around." These arrangements further ameliorate the risk that appellant will suffer financial harm if its customers use unlawfully competing cotton marketing software. The district court's finding that appellant failed adequately to demonstrate irreparable injury is not clearly erroneous.

Finally, these findings by the district court are not incorrect as a matter of law. Appellant attempts to buttress his argument that copying occurred in fact by the assertion that copying occurred as a matter of law even if appellees relied solely on their memory to construct their new cotton marketing program. Appellant cites cases from Third, Seventh, and Ninth Circuits to establish the proposition that copying done from memory is still copying, and is no less objectionable than any other form of copying.3 These cases do not, however, rebut the conclusion of the district court that copying of protected expression had not been sufficiently demonstrated.

The legal finding by the district court ultimately rests on a judgment about the extent of the protection offered by appellant's copyright. On that issue, we look to our colleague Judge Higginbotham's opinion in Synercom Technology, Inc. v. University Computing Co., 462 F. Supp. 1003 (N.D.Tex.1978). In that case, Judge Higginbotham held that "input formats" of a computer program--the organization and configuration of the information fed to the computer--were ideas, not expressions, and thus were not protected by copyright. 462 F. Supp. at 1014.

To the extent that input formats represent a level of computer software design more specific than functional design and more general than line-by-line program design, the issue of their copyrightability is relevant to the issue of whether GEMS infringes on protected Telcot designs. Appellant urges that we adopt the reasoning of Whelan Associates, Inc. v. Jaslow Dental Laboratory, Inc., 797 F.2d 1222 (3rd Cir.1986), which is admittedly "at odds with Judge Higginbotham's scholarly opinion." Id. at 1239. Whelan rejects the premise developed in Synercom that "there [is] a difference between the copyrightability of sequence and form in the computer context and in any other context," id. at 1240, holding that the structure, sequence, and organization of computer programs are copyrightable.

We decline to embrace Whelan for two reasons. First, the issue is presented to us on review of a denial of a motion for preliminary injunction. Thus, the record is only partially developed, and our review is one step removed from the actual merits of the case. Second, appellees presented evidence that many of the similarities between the GEMS and Telcot programs are dictated by the externalities of the cotton market.4 To that extent, the facts of this case fit squarely within Synercom's powerful analogy to the hypothetical development of gear stick patterns. 462 F. Supp. at 1013. The record supports the inference that market factors play a significant role in determining the sequence and organization of cotton marketing software, and we decline to hold that those patterns cannot constitute "ideas" in a computer context.

Appellant has thus failed to demonstrate that the factual findings of the district court regarding the copyright infringement claim are clearly erroneous or that the findings of law are incorrect. We affirm the denial of the motion for a preliminary injunction on the copyright claim.

Appellant's request for a preliminary injunction was also based on its claim that appellees improperly made use of its trade secrets. Misappropriation of trade secrets is shown by proof: (1) that a trade secret existed, (2) that the trade secret was acquired through a confidential relationship, and (3) that the defendant used the trade secret without authorization from the plaintiff. Black, Sivalls & Bryson, Inc. v. Keystone Steel Fabrication, 584 F.2d 946, 951 (10th Cir.1978), citing Official Airlines Schedule Information Service v. Eastern Air Lines, 333 F.2d 672, 673-74 (5th Cir.1964). The district court found that appellant could not demonstrate a sufficient likelihood of success on the merits because it failed to satisfy the third prong of this three-point test.5

Admittedly, the district court's findings do not differentiate between the copyright and trade secrets claims. The record shows, however, that the court did consider the trade secret claim alongside the copyright claim throughout. It is true that a protected interest in trade secrets is more easily created than such an interest in copyrighted material. But the trade secrets allegedly involved here are particular implementations of software functions. The critical point is that the misuse of these implementations can occur only through copying the particular software designs on a sufficiently specific level. These are not alleged marketing or mode of doing business trade secrets. They are matters of design, where the issue of misuse boils down to evidence of copying. If no copying occurred on any level, appellant cannot demonstrate that appellees misused the trade secrets they allegedly possessed.

The district court relied in its findings upon the opinion of appellees' expert witness that no copying of appellant's software occurred. We have discussed above the evidence in the record supporting this finding. Considering the evidence presented, the trade secrets battle was nothing more than a restaging of the copyright battle. If appellees did not in any way "copy" any part of appellant's protected idea or expression, then appellant cannot demonstrate trade secret misappropriation any more than it can show copyright infringement.

The district court's finding is also supportable as a conclusion of law. While an employee owes a duty to his employer not to disclose any knowledge of that employer's trade secrets, the "employee, upon terminating his employment relationship with his employer, is entitled to take with him 'the experience, knowledge, memory, and skill, which he ... gained while there employed.' " Anaconda Co. v. Metric Tool & Die Co., 485 F. Supp. 410, 423 (E.D.Pa.1980) (citation omitted). Also persuasive is the statement in Structural Dynamics Research Corp. v. Engineering Mechanics Research Corp., 401 F. Supp. 1102, 1111 (E.D.Mich.1975): "[I]f the subject matter of the trade secret is brought into being because of the initiative of the employee in its creation, innovation or development even though the relationship is one of confidence, no duty arises since the employee may then have an interest in the subject matter at least equal to that of his employer or in any event, such knowledge is a part of the employee's skill and experience."

Whether or not the development of GEMS involved the application of appellees' skill and experience or of appellant's trade secrets is a question of fact. The district court's factual finding, then, distinguishes this case from those relied upon by appellant, where the defendant did make use of trade secrets. See, e.g., FMC Corp. v. VARCO International, Inc., 677 F.2d 500 (5th Cir.1982); Weed Eater, Inc. v. Dowling, 562 S.W.2d 898 (Tex.Civ.App.1978, writ ref'd n.r.e.); Electronic Data Systems Corp. v. Powell, 524 S.W.2d 393 (Tex.Civ.App.1975, writ ref'd n.r.e.).

Finally, appellant argues that the district court violated Fed.R.Civ.P. 52(a) by failing adequately in terms to address appellant's trade secret misappropriation claim. Rule 52(a) requires that "in granting or refusing interlocutory injunctions the court shall similarly set forth the findings of fact and conclusions of law which constitute the grounds of the action." We find that the district court discharged its duty under this rule. As explained above, the record supports the tentative determination that no copying occurred. The court's statement that appellant failed to demonstrate a likelihood of success on the merits is therefore as applicable to the trade secrets claim as it is to the copyright infringement claim, and satisfies the court's obligation under Fed.R.Civ.P. 52(a).

Appellant has failed to demonstrate that the factual findings of the district court regarding the trade secrets misappropriation claim are clearly erroneous or that the findings of law are incorrect. We affirm the denial of the motion for a preliminary injunction on the trade secrets claim.

Appellant failed to demonstrate a sufficient likelihood of success on the merits of its claims, and failed to present sufficient proof of irreparable harm to justify a preliminary injunction. The judgment of the district court denying the motion for a preliminary injunction is

AFFIRMED.