Jostens, Inc. v. National Computer System, Inc.

318 N.W.2d 691 (1982)

JOSTENS, INC., Appellant,
v.
NATIONAL COMPUTER SYSTEMS, INC., et al., Respondents.

No. 81-637.

Supreme Court of Minnesota.

April 23, 1982.

Oppenheimer, Wolff, Foster, Shepard & Donnelly, Elmer B. Trousdale, Ronald C. Mickelberg and Paula D. Osborn, St. Paul, Orville E. Fisher, Jr., Gen. Counsel, Minneapolis, for appellant.

Leonard, Street & Deinard, Harold D. Field, Jr., Stephen J. Davidson and Robert L. DeMay, Minneapolis, for respondents.

Heard, considered and decided by the court en banc.

SIMONETT, Justice.

Appellant, Jostens, Inc., sued several of its former employees and a computer firm for misappropriation of trade secrets and proprietary data in Jostens' computer system used for the design and manufacture of its class ring molds. The trial court found no misappropriation. Because the trial court's findings have support in the evidence, we affirm.

This appeal involves application of the trade secret doctrine to the rapidly expanding and highly complex field of computer technology.^[1] The appellant-plaintiff is Jostens, Inc., a Minnesota corporation, which manufactures and markets school products, including student class rings. Jostens sued National Computer Systems, Inc. (NCS), a Minnesota corporation which designs and makes computer systems;^[2] John S. Titus, Jr., once an employee of Jostens and now an officer of NCS; and Robert J. Henderson and Allan Hoagberg, also former employees of Jostens who joined NCS.

For many years Jostens had used skilled artisans to engrave designs into the molds for the shanks of its rings. John Titus, an engineer at Jostens, became interested in developing a computer system to be used in making the ring molds. In 1972 and 1973, Titus began putting such a system in Jostens' plant at Burnsville. By early 1974 the system was in operation and operating successfully, giving Jostens a competitive advantage over other ring manufacturers still using traditional manual methods of producing ring shank molds. In about January 1978, Jostens lost its position as the sole class ring manufacturer with a computerized mold-making system, when respondent National Computer Systems, Inc., designed and sold a similar computer system to L. C. Balfour Company, a competing ring manufacturer. That sale triggered this lawsuit.

The computer system before us is called a CAD/CAM system, which stands for a computer-aided design and computer-aided manufacturing system. Jostens' CAD/CAM system consisted of three subsystems: (1) a digitizer or scanner subsystem, which translates positional data from artwork and three-dimensional models into computer-readable magnetic tape, which is then fed into, (2) the interactive computer graphics subsystem, in which an image is displayed in three dimensions on a screen where it can be manipulated and corrected by an operator; and (3) the engraving subsystem, where the computer, instructed by data from a magnetic tape, in a process called numeric control, guides a machine which engraves the design on the mold for the ring shanks.

Titus purchased each subsystem from a different vendor. The scanner subsystem was purchased from Potter Instrument Company; the graphics subsystem was purchased from Adage, Inc., of Boston; and the hardware components of the engraving subsystem, such as the engraving machine and the controller, were purchased from different vendors and then connected at the Burnsville plant by Titus and the vendors.

Most of the focus at the trial was on the graphics subsystem purchased from Adage. It consisted of both hardware and software. The hardware was a standard item; the software, i.e., the instructions to the computer, was of two kinds — operations systems software and application software.^[3] Of these two types, the application software was written by Adage pursuant to Jostens' functional specifications and sold to Jostens for $49,500. Jostens claims the overall cost to put in its CAD/CAM system was the largest authorization for capital expenditure in the company's history. When Jostens placed its order with Adage for the graphics system, it added the following clause: "Proprietary — All materials propared [sic] for Jostens' specific requirements shall become the property of Jostens."

In the midseventies, after Jostens' system was successfully in use, Titus urged his employer to consider merchandising the CAD/CAM system by selling either service or equipment to other manufacturers. Some interested persons were invited to tour the Burnsville plant, but in May 1975, Jostens decided against any further efforts to commercialize the system and ordered the Burnsville facility closed to outsiders. Also at that time, Titus, with Jostens' permission, spoke at a conference of the Numerical Control Society in Washington, D. C., about Jostens' CAD/CAM system, giving a multimedia presentation followed by publication of an article in a technical journal, also authorized by Jostens, which included a description of the major components of Jostens' system and the system's role in the production of ring molds.

In August of 1975, Titus left Jostens and soon thereafter began working full time for NCS, whose chief executive was Charles Oswald, a former president of Jostens. Apparently Titus' departure was amicable. Jostens knew that, in his new job at NCS, Titus would be working on the development of CAD/CAM technology for the general market. As Jostens' president put it, "Exactly how much that was exactly like Jostens' system, specialized tooling, I was not aware and was not concerned because I think we wanted John to go ahead and work in this area. It was fine with us." On the other hand, Titus had signed an agreement in 1969, 4 years after he was hired by Jostens, acknowledging that all papers prepared by him were Jostens' property and that he would not reveal to others any information concerning Jostens' business "including its inventions, shop practices, processes and methods of manufacturing and merchandising."

Soon after Titus began work at NCS, his new firm proposed a technical cooperation agreement to Jostens under which the hardware and software developed by NCS would be kept compatible with Jostens' Burnsville system in order to provide for current backup and future improvements. To ensure compatibility, the proposal provided that Jostens would furnish NCS with a current copy of its application software package and NCS would promise to refrain from marketing CAD/CAM systems to any of Jostens' competitors. During the same period that NCS and Jostens were considering the cooperation agreement, NCS was also negotiating with Adage, the vendor that had furnished Jostens with its graphics subsystem. NCS wanted Adage to supply it with a software package for its own CAD/CAM system, and it appears that NCS was hoping to get a copy of the software package of Jostens' graphics subsystem for Adage to use in this new project, at least as a reference point for the broader, basic system that NCS wished to develop. In February 1976, Jostens rejected the proposed technical cooperation agreement.

Although collapse of the proposed agreement meant that NCS was unable to get a copy of Jostens' software package, NCS continued plans for the development of a CAD/CAM system, still working with Adage. Titus says that he was told by Adage that NCS' proposed system would be "considerably more difficult" than Jostens' and that there would be "very little, if any, usability of Jostens' software per se." Oswald says he was told by Adage that they had checked with their own attorneys about whether they could produce the software for NCS notwithstanding their earlier work on Jostens' software and that counsel had said, "[I]t's no problem."

A central issue at trial was what and how much of the Adage material used to write Jostens' original package was also used to write NCS' programs. Specifically, this involved not the operation systems software, a standard component sold to all Adage customers, but the application software, used to adapt a system to a particular user's needs. Adage's programmers testified many of the routines they used in writing NCS' package were "utility routines," simply taken off their library shelf, and that in assembling new application packages, programmers usually wrote only about 10% new material, while in this case about half of the final NCS package was original work. Before Adage had completed the NCS software, Adage decided to discontinue its application software business. NCS then continued development of the software on its own and argued at trial that the version it eventually sold had been significantly changed and improved from that which NCS had received from Adage.

In the summer of 1977, NCS approached Balfour, a class ring manufacturer and competitor of Jostens, about the sale of a CAD/CAM system; by late September 1977, Balfour had tentatively decided to purchase NCS' model. One of the sale terms discussed was Balfour's request that NCS indemnify Balfour against any possible trade secret claim, specifically a trade secret challenge Jostens might make. Balfour says it was not aware of any facts to support such a claim but it was wary since it knew NCS' president, Oswald, had once been Jostens' president. Ultimately, the two parties agreed that NCS would repurchase the system from Balfour on a depreciated basis in the event of a suit by Jostens against Balfour within 5 years. Also in late 1977, while NCS and Balfour were negotiating a sale, NCS sent a letter to Jostens urging it to reconsider a decision it had made not to buy equipment from NCS. In the letter NCS said there was a large market for its new CAD/CAM system and "[w]e plan these installations to be made in the jewelry, class ring and toy industries." Jostens neither responded nor objected to NCS' apparent intent to sell its system on the market.

The sale and delivery of NCS' CAD/CAM system to Balfour took place in the period of December 1977 to February 1978. It appears that earlier, sometime between March and September 1976, Jostens' vice president of manufacturing had become aware that NCS might be using software programs developed by Adage for Jostens, and voiced concern over this to Jostens' vice president and general manager of the scholastic division. Defendants also contend Jostens should have understood as early as February 1976, when Jostens rejected the technical cooperation proposal which included an explicit noncompete agreement, that NCS might eventually market a system to other ring manufacturers. Jostens, however, did not start its lawsuit until July 1978. Thus defendants also raised issues of estoppel and laches, both of which the trial court found to be present.

As damages for misappropriation of its trade secrets, Jostens claimed all of NCS' gross profits on its CAD/CAM system and, from the three individual defendants, sums totaling salaries, stock option benefits, and commissions. Jostens also claimed a royalty of NCS revenues, punitive damages, and a permanent injunction prohibiting the defendants, among other things, from using, selling or leasing any part of any CAD/CAM system. The trial court denied all of these requests. For their part, defendants requested an award of their disbursements, a request also denied by the trial court, for which defendants petition for review here.

ISSUES

The case presents these broad issues: (1) Did Jostens possess a trade secret and, if so, of what did it consist; (2) as a corollary, was there any misappropriation of a trade secret by the defendants; (3) did Titus or Henderson breach their employment agreements with Jostens by revealing and using Jostens' business data; (4) is Jostens estopped or barred by laches from asserting its claims; and (5) is NCS entitled to its disbursements?

Since the trial court found no trade secret, no misappropriation and no breach and also found estoppel and laches, these issues are reviewed here in the form of whether the trial court's findings of fact are clearly erroneous. Minn.R.Civ.P. 52.01. The trial was long and complex. The parties supplemented their testimony (including the live testimony of seven experts) with film, slide and videotape presentations; in addition, the trial judge visited both Jostens' and NCS' plants. The trial judge's decision includes detailed findings and an extensive memo. Here, we can give space only to some of the salient facts.

I. The existence of a trade secret.

It is not always easy to follow Jostens' contentions because its claim of a trade secret is rather elastic. At times, the claim appears to include the entire CAD/CAM system; at other times, something less.

Plainly, Jostens is claiming secret status for the computer graphics subsystem, purchased from Adage, and for the customizing work done at Jostens' own plant to connect the subsystem and make the adjustments needed for the manufacture of ring molds. This seems to be the clearest and most plausible of Jostens' claims, but even here clarity is not always present. Does the trade secret encompass all three parts of Adage's graphics subsystem? For there is the hardware (a standard feature of Adage packages), the operating system software (also standard), and the application software (a more particularized feature, using some standard or utility routines). Or does the claim include only those programs and routines within the application software segment specifically written for Jostens? Or does the claim lie in the distinct combination of all these parts, standard and original, in Jostens' system?

A. Before reaching these questions, we must first consider a preliminary issue: What is the effect of the proprietary clause in Jostens' purchase order to Adage, where it is said, "all material prepared for Jostens specific requirements shall become the property of Jostens"? The trial court held this clause was never accepted by Adage and did not become a contract provision. We agree.

1. Adage's own quotation stated that the contract did not include research or development work by Adage and, in addition, expressly provided that no variation from the terms of the quotation would become a part of the contract unless specifically approved in writing by Adage. It is clear Adage never specifically approved Jostens' proposed proprietary clause. Mere performance and delivery of the items ordered, together with silence concerning the proposed additional clause, did not constitute an acceptance of that term where acceptance was expressly conditioned on the seller's assent to additional terms. Uniroyal, Inc. v. Chambers Gasket & Manufacturing Co., 380 N.E.2d 571 (Ind.Ct.App.1978); see Minn.Stat. § 336.2-207(2)(a) (1980).

2. The trial court found, however, that the parties, nevertheless, did have an understanding that Adage would not sell to any other party a duplicate of Jostens' application software package and that, in the light of industry custom and practice, "Jostens' sole proprietary interest was in the application software package as a whole, and not in the individual routines or lines of code." See Minn.Stat. §§ 336.2-207(3), 336.2-208(2) (1980). There was evidence to support this finding and we cannot say it is clearly erroneous.

To say, however, that Jostens had a proprietary interest in the Adage application software package "as a whole," does not advance Jostens' claim very far. No one claims Adage sold a duplicate package to anyone else, not even to NCS. No one is suing Adage. On the other hand, Jostens' proprietary interest in the "whole" package may have a bearing on defendant Titus' conduct, reflecting on what he believed to be his employer Jostens' property.

B. This brings us, then, to the main issue, whether Jostens had a trade secret in the CAD/CAM system or any of its parts or any combination of the parts. The trial court held Jostens did not, and we agree.

1. In Cherne Industrial, Inc. v. Grounds & Associates, Inc., 278 N.W.2d 81, 90 (Minn. 1979), we held that a trade secret has four characteristics: (1) the matter involved is not generally known or readily ascertainable, (2) it provides a demonstrable competitive advantage, (3) it was gained at expense to the plaintiff-owner, and (4) the plaintiff-owner intended to keep it confidential. This characterization has been given statutory recognition under our state's Uniform Trade Secrets Act, enacted in 1980, after the events of this lawsuit.^[4]

Here, however, the trial court found that Jostens failed to establish any of the four Cherne requirements. Although it appears that appellant's proof fell short on the second and third Cherne requirements, we need not decide this, as we find only the first and fourth need to be discussed.

2. The first requirement is that the information be not generally known or readily ascertainable. Courts agree that trade secrets lie somewhere on a continuum from what is generally known in a field to what has some degree of uniqueness, although there need not be the degree of novelty or originality required for patent or copyright protection. Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 476, 94 S. Ct. 1879, 1883, 40 L. Ed. 2d 315 (1974). Within these limits, courts have suggested a variety of further limitations. Some measure of discovery is required. Koehring Co. v. E. D. Etnyre & Co., Inc., 254 F. Supp. 334, 361 (N.D.Ill.1966). Mere variations in general processes known in the field which embody no superior advances are not protected. American Chain & Cable Co., Inc. v. Avery, 143 U.S.P.Q. 126, 130 (1964). But unique principles, engineering, logic and coherence in computer software may be accorded trade secret status. Com-Share, Inc. v. Computer Complex, Inc., 338 F. Supp. 1229, 1234 (E.D.Mich.1971). And a trade secret may modify and improve standard models to a point at which the newer version is unique in the industry. Kubik, Inc. v. Hull, 56 Mich.App. 335, 224 N.W.2d 80 (1974).

Further, generally known computer elements may gain trade secret protection from the nature of their combination. Cybertek Computer Products, Inc. v. Whitfield, 203 U.S.P.Q. 1020 (1977). Thus a combination of elements produced a unique and valuable computer software program in Structural Dynamics Research Corp. v. Engineering Mechanics Research Corp., 401 F. Supp. 1102 (E.D.Mich.1975). See also Allen Manufacturing Co. v. Loika, 145 Conn. 509, 144 A.2d 306 (1958).

3. Jostens starts with the fact it had the first and only CAD/CAM system in the jewelry ring industry. In that sense, perhaps, it can be said this system was not generally known; but, as the trial court found, the technology involved in CAD/CAM systems was both generally known and readily ascertainable. There was evidence that computer aided graphics systems were used for machine tooling in industries other than ring manufacture before Jostens ordered its CAD/CAM system and that the concept was known to the industry by the early sixties and is still developing. A defense witness and computer graphics expert involved in the design of similar hardware and software for Bell Telephone testified that within his company design teams had worked on interactive graphics programs performing functions similar to those performed by the Jostens system. Both the scanner subsystem and the engraving subsystem as well as the hardware and operating systems software for the graphics subsystem were all standard vendor products.

4. Although Jostens had purchased the hardware from outside vendors, it argues what is important is that it "built a system by combining components from a number of vendors because no complete system was commercially available." It argues substantial "customization" was needed to make the commercially available components into a productive system, and it points to such instances as Titus' work in constructing a three-axis capability for the engraving table.

The trial court found, however, "that the assembly of Jostens' CAD/CAM system did not require substantial research or experimentation" and that the system came about "through Titus' application of his general skill and knowledge to the integration of commonly available components to perform the desired function." We cannot say, after reviewing the evidence, that the trial court's findings are clearly erroneous. Clearly, the CAD/CAM system as such, as the combination of three generally known subsystems, does not achieve the degree of novelty or "unknownness" needed for a trade secret. As to a combination of lower levels within the system or within a subsystem, the likelihood of novelty increases, but even here we do not find plaintiff's burden of proof has been sustained. Again, we are plagued with the elasticity of plaintiff's claim. Simply to assert a trade secret resides in some combination of otherwise known data is not sufficient, as the combination itself must be delineated with some particularity in establishing its trade secret status. Cf. Struthers Scientific & International Corp. v. General Foods Corp., 51 F.R.D. 149, 153, 168 U.S.P.Q. 202, 204-05 (D.Del.1970) (requiring greater specificity in interrogatories regarding alleged trade secret in combination of elements).

5. Beyond a trade secret claim based on combination of parts or customization, Jostens spends the most time urging that the application software portion of the graphics subsystem, prepared for it by Adage, was not generally known or reasonably ascertainable. Jostens points out that when it placed its order with Adage, Adage at that time had never done an industrial numerical control application. On the other hand, there was evidence that Jostens' application software package was assembled by Adage's modified use of two application software systems (ORTHO and Cubic) that it already had in hand, having been written for other prior customers. Adage had used a modular or structured program practice, putting together small, self-contained routines and using them as building blocks for new application packages. Like others in the industry, Adage maintained a library of previously written routines and programs to use in building new programs. Experts for the defendants testified Jostens' application software did not involve any new or innovative advances in algorithmic technique. On the conflicting evidence the trial court could find, as it did, "that Jostens' CAD/CAM system did not represent an `invention', a discovery of any kind, or a novel technological contribution which differed materially from methods already well known in the field of manufacturing engineering, and that * * * Jostens' system was no different in concept from other systems already in the public domain."

6. Even if a trade secret were not generally known or readily ascertainable, it might become so if its possessor disclosed it to the public. Here, especially, Jostens' proof fails to persuade. The trial court found that Jostens failed to meet the fourth requirement of the Cherne test, namely, that it intended to keep the relevant information secret. We agree.^[5]

Secrecy need not be total; depending on the circumstances, only partial or qualified secrecy will do. Radium Remedies Co. v. Weiss, 173 Minn. 342, 217 N.W. 339 (1928). The information must be, as Minn.Stat. § 325C.01, subd. 5 (1980), now puts it, "the subject of efforts that are reasonable under the circumstances to maintain secrecy." See also 1 Milgrim, Trade Secrets § 2.04 (1980). These efforts may extend both to internal secrecy, keeping the information in-house, and to external secrecy, keeping the information from those outside in the general trade or industry. Thus, in Com-Share, Inc. v. Computer Complex, Inc., 338 F. Supp. 1229 (E.D.Mich.1971), computer software was protected both by plaintiff marking each page of the listings to emphasize its system as "confidential" and by building passwords into the system to prevent unauthorized access. Employees need to understand that information which is not readily available to the trade is not to be made so by them. Kubik, Inc. v. Hull, 56 Mich.App. 335, 358, 224 N.W.2d 80, 92 (1974). The plaintiff employer in Structural Dynamics Research Corp. v. Engineering Mechanics Research Corp., 401 F. Supp. 1102 (E.D.Mich.1975), specifically called the confidential nature of the work to each employee's attention in an individual confidential disclosure agreement each signed. On the other hand, in Pressure Science, Inc. v. Kramer, 413 F. Supp. 618 (D.Conn.1976), the court found that a plaintiff's failure to require all employees working in a supposedly confidential area to sign a nondisclosure agreement evidenced a fatal lack of concern for confidentiality.

Here there was evidence that when Jostens installed its CAD/CAM system, no consideration was given or policy established to keep the development secret or confidential. Not until May 1975 did Jostens bar potential customers for its system from the Burnsville plant; until then, prospects were allowed in, although it is disputed what they might have learned. Particularly damaging, we think, is the presentation made and the article written by Titus, with Jostens' approval, explaining Jostens' CAD/CAM system to other experts in the field. The parties disagree on whether the information revealed was sufficient to make the purported trade secret accessible to other technologically sophisticated persons. NCS' expert testified that from Titus' article alone he could have duplicated the functions of Jostens' system, while an expert called by Jostens said it would be "very, very difficult" to translate into practical application the theoretical concepts described in the article. The trial court found that the information given in Titus' authorized disclosure was sufficient to enable an experienced engineer to duplicate both the hardware and the software of Jostens' system "without too much difficulty."

There was also evidence that none of the software tapes or documents at Jostens was marked "secret" or "confidential" until after this litigation began. Some of the employees working with the system were never asked to sign a confidentiality agreement even though Jostens had employed such a form elsewhere.

We hold the trial court's finding that "Jostens did not take reasonable steps to protect its alleged trade secrets" is not clearly erroneous.

II. The claim of misappropriation.

This case may also be viewed from the standpoint of whether there was a misappropriation of a trade secret. This analysis centers on the misappropriation, the gist of which is breach of a confidential relationship, rather than on whether a property interest is involved, i.e., whether a trade secret exists.^[6] The protection afforded trade secrets is not intended to reward or promote the development of secret processes (although it does, of course, benefit the enterprising developer), but rather is to protect against breaches of faith and the use of improper methods to obtain information. 1 Milgrim, Trade Secrets § 12.01. Trade secret law seeks to maintain standards of loyalty and trust in the business community.

A. These two approaches overlap, since one of the requirements of a misappropriation cause of action is to first prove the existence of a trade secret. Plaintiff must then further prove that defendant acquired the trade secret as a result of a confidential relationship and that the defendant has used and disclosed the trade secret. These three elements are set out in Eutectic Welding Alloys Corp. v. West, 281 Minn. 13, 18, 160 N.W.2d 566, 570 (1968). What needs to be kept in mind, however, is that the various elements "should not be artificially separated for purposes of analysis since, in a significant sense, they are interdependent." 1 Milgrim, Trade Secrets § 7.07(1) at 95.

Here, Jostens contends that defendants Titus, Henderson and Hoagberg acquired knowledge of Jostens' trade secrets while in its employ, and they then used this information to produce a similar CAD/CAM system for their new employer, defendant NCS. As to Henderson, for example, the claim is that he used his experience gained at Jostens to train personnel at Balfour in the operation of the CAD/CAM system sold it by NCS. As to Titus, the claim is that he took his knowledge, experience and skills to NCS and used them to build a similar system for NCS. Defendants' defense to the claim of misappropriation was that there were no trade secrets, nothing to misappropriate, and in this connection that (1) NCS developed its system using only generally known computer concepts which are not protectable, and (2) if any of the concepts might be protectable, the protection was lost by Jostens' failure to take reasonable precautions against disclosure and use, and (3) the defendants developed their product independently from other sources.

B. First of all, even without the employment agreements signed by Titus and Henderson, employees have a commonlaw duty not to wrongfully use confidential information or trade secrets obtained from an employer. We do not find this commonlaw duty to have been breached.

1. Titus took with him (how could he not?) his experience and skills acquired while working for Jostens.^[7] These abilities, to the extent derived from generally known sources, are not considered confidential; a computer programmer, like a real estate salesperson, should be able to ply his trade. Jim W. Miller Construction, Inc. v. Schaefer, 298 N.W.2d 455, 459 (Minn. 1980). But knowledge gained at an employer's expense, which takes on the characteristics of a trade secret and which would be unfair for the employee to use elsewhere, is deemed confidential and is not to be disclosed or used. Cherne Industrial, Inc. v. Grounds & Associates, Inc., 278 N.W.2d 81 (Minn.1979). Even if this knowledge is only in the employee's memory, it may be protectable. Cybertek Computer Products, Inc. v. Whitfield, 203 U.S.P.Q. 1020 (1977). Confidential information is that which an employee knew or should have known was confidential. Structural Dynamics Research Corp. v. Engineering Mechanics Research Corp., 401 F. Supp. 1102 (E.D.Mich. 1975). Or as Minn.Stat. § 325C.01, subd. 3(ii)(B)(II) (1980), now puts it, a person, such as an employee here, is not to disclose or use a trade secret that he knows or has reason to know was acquired by him under circumstances giving rise to a duty to maintain its secrecy or limit its use. On the other hand, the employee is entitled to fair notice of the confidential nature of the relationship and what material is to be kept confidential. Ellis, Trade Secrets 79 (1953).

2. Jostens' failure to take reasonable precautions to protect the confidentiality of what it now claims to be secret was such that the defendant employees could not be expected to have known what was confidential and what was not, what was unfair to disclose and what was not. The trial court found, and there is evidence to support the finding, that Jostens did not really know what it was about its Burnsville system that it intended to protect. Perhaps this difficulty existed because of all people at Jostens Titus, the person against whom protection is now sought, was the most knowledgeable about what information should be protected. It appears Titus was not unaware of his position. Titus recognized Jostens' "proprietary interest" in his project. Thus he sought and obtained his employer's permission to disclose aspects of the CAD/CAM to the trade in his seminar presentation and journal article. Still, it can hardly be said that Jostens had an understanding with Titus not to disclose any particular information when, in lieu of any "exit interview," Jostens' president, on Titus' departure, said the company wanted Titus to go ahead and work in the area of CAD/CAM development and this "was fine with us."

Thus the trial court was not in error in finding that at the time Titus left Jostens his knowledge of the operation of the Burnsville system was "part of his own general skill and experience." If some of Titus' knowledge might nevertheless be a likely subject for a trade secret, here, we think, Jostens failed to make it so by failing to insist on its secrecy. This is true also for the other defendant employees.

3. A misappropriation claim also requires proof under Eutectic that the defendants disclosed or used the trade secret. Much evidence was received on this issue. Jostens tried to show that the application software package used in the NCS graphics subsystem contained significant parts of the same software package in Jostens' system. Jostens showed, for example, through a comparison study made by two of its experts, that some 29 names of routines were identical and, within routines bearing the same names, a high percentage of the lines were syntactically identical.^[8] Jostens showed that copies of some of the source code furnished NCS by Adage were labeled "Jostens Application Software." On the other hand, Adage's programmer testified that only a small percentage of NCS' software routines, at most 10%, had been written from Jostens' package and that while a somewhat higher percentage had begun as routines for the Jostens package, they were substantially revised before being used in the NCS package. Then, too, some of the so-called Jostens routines were identical as standard utility routines, kept in Adage's library and used as building blocks as needed.

The fact that only a small percentage of Jostens' routines may have been used does not especially help the defendants; it may take only that small percentage, depending on what it entails, to make a program unique. Nevertheless, Jostens did not convince the trial court that NCS' system itself evidenced use of any of Jostens' alleged trade secrets. The court made extensive findings of specific ways in which the NCS system had functional capabilities totally nonexistent in or substantially different from Jostens' software package. There was testimony by an Adage programmer that at the time the technical cooperation agreement was still a possibility, he had already determined that NCS' proposed program would require such enormous modification of Jostens' package that Adage could as well start all over anyway. In light of such evidence, we cannot say that the trial court's findings that Jostens failed to prove disclosure or use of its claimed trade secret were clearly erroneous.

C. A confidence may be imposed contractually as well as by employment status, so Jostens also asserts a separate claim that defendants Titus and Henderson are liable for breach of their employment agreements. The trial court found the agreements unenforceable on the grounds neither was supported by consideration. We agree.

1. Four years after Titus began his employment with Jostens and two years after Henderson began his, they each signed the following agreement:

All papers and apparatus relating to Jostens' business, including those prepared or made by me, shall be the property of Jostens and except as required by my work, I will not reveal them to others nor will I reveal any information concerning Jostens' business including its inventions, shop practices, processes and methods of manufacturing and merchandising.

There was no evidence that, by signing this agreement, either Titus or Henderson gained greater wages or a promotion or access to technical or operational parts of the Burnsville system that nonsigning employees did not have. According to Titus, he and the others asked to sign the agreement did so under the impression that they would lose their jobs if they did not. Jostens argues that the promise of future or continued employment is adequate consideration.

2. We have held that the adequacy of consideration for a noncompetition contract in an ongoing employment relationship depends on the particular facts of each case. Davies & Davies Agency, Inc. v. Davies, 298 N.W.2d 127 (Minn.1980). In Davies we observed that the contract provided the employee with real advantages. That does not appear to be the case here. The agreements obtained by Jostens did not increase Jostens' commitment to the employees for future benefits, as in Modern Controls, Inc. v. Andreadakis, 578 F.2d 1264 (8th Cir. 1978) (applying Minnesota law). Nor do the agreements memorialize a prior oral agreement made between the employer and prospective employee, as in Cybertek Computer Products, Inc. v. Whitfield, 203 U.S.P.Q. 1020 (1972).

Davies contains dicta that "[m]ere continuation of employment as consideration could be used to uphold coercive agreements." Id., 298 N.W.2d at 130-31 (emphasis added). In this instance, however, where no raises or promotions resulted, where other employees with similar access were not asked to sign, the mere continuation of employment for Titus and Henderson is not enough. To the extent the agreements encompassed the employees' common-law duty to their employer not to disclose, our disposition of that common-law duty makes any contractual duty moot. Since we find the employment agreements not enforceable, we need not decide the issue of whether their purported reach of "all" papers and information is overbroad or lacking in fair notice of what is not to be disclosed.

III. Other claims.

Since we uphold the trial court's findings that there has been no misappropriation of any trade secret, we need not reach the issues of whether the trial court's findings of estoppel and laches are sustainable.

Respondents petition for review of that portion of the trial court's decision denying them an award for their disbursements. "In every action in district court," says Minn.Stat. § 549.04 (1980), "the prevailing party * * * shall be allowed disbursements necessarily paid or incurred" (emphasis added). Here respondents, who had asserted no counterclaim, are clearly the prevailing party and entitled, by statute, to their taxable disbursements. We remand, therefore, to amend the decision in this respect and for disbursements to be taxed.

Affirmed, but with respondents to be allowed their disbursements.

NOTES

[1] For example, it is estimated that some 15,000 computer programs are written each day in this country and that the total value of this software is in the tens of billions of dollars. See W. Schmidt, Legal Proprietary Interests in Computer Programs: The American Experience, 21 Jurimetrics Journal 345 (1981) (hereinafter referred to as Schmidt).

[2] Also sued is Data Input, Inc., a local corporation which was a subsidiary of NCS from 1972 to 1976 and was then merged into NCS. For simplicity, we will consider Data Input, Inc.'s role as subsumed in NCS' conduct.

[3] "Computer software tells computer hardware to generate infinitesimally brief, timed pulses of electricity according to a programmed sequence, hence the word program. Computers operate by sensing the presence or absence of these electrical charges at each cycle of their operation. The presence or absence of such charge is represented by a `1' or a `0,' respectively, in written program representations called object code." Schmidt at 351. Computers understand only these pulses. So, as Schmidt explains, "source languages" were invented to enable a person to communicate with the machine, to be able to write a program. A source language is then rewritten into "assembly language," consisting of brief mnemonics as a kind of bridge between human and computer language.

The operations system software is usually considered an integral part of the hardware provided by the manufacturer and consists of the basic programming necessary to enable the machine to be operated, i.e., to be applied by further programming to the task to be performed. Application software, on the other hand, consists of these further programs which enable the machine to apply itself to the task set for it.

Programs, both in their intangible sense and as embodied in tapes, disks, and instruction manuals, may be considered software.

[4] Minn.Stat. § 325C.01, subd. 5 (1980), provides:

"Trade secret" means information, including a formula, pattern, compilation, program, device, method, technique, or process, that:

(i) derives independent economic value, actual or potential, from not being generally known to, and not being readily ascertainable by proper means by, other persons who can obtain economic value from its disclosure or use, and

(ii) is the subject of efforts that are reasonable under the circumstances to maintain its secrecy.

It appears that the statutory definition of a trade secret does not require Cherne's third element, that the information was gained at expense to the owner. See also Restatement of Torts § 757 (1939).

[5] The possessor of a computer program trade secret may seek copyright or patent protection but these protective measures have their own drawbacks. See Bender, Trade Secret Protection of Software, 38 Geo.Wash.L.Rev. 909 (July 1970). Thus the possessor of a trade secret, especially when there is no intention to market the product, may choose, as Jostens apparently has done here, to seek instead trade secret protection.

[6] Whether a trade secret is a form of property usually arises in the context of the owner's attempts to protect the secret and the court's attempt to fashion appropriate relief. But as Milgrim points out, "the property view underlies protection of trade secret decisions and is, in fact, the keystone upon which the protection body of case law rests." 1 Milgrim, Trade Secrets § 1.09.

[7] Just before the lawsuit was started, Titus furnished Jostens with his affidavit that he took with him only some insurance and stock option information and a "file relating to CAD/CAM which had names of manufacturers and contacts for the hardware, because I intended to order some of the equipment to continue my work on CAD/CAM." It developed, however, that Titus had kept some documents which one of Jostens' experts testified might be helpful in duplicating Jostens' CAD/CAM system. Titus said he made no use of the documents. The trial court found that "Titus did not retain any documents which he thought might be within [the scope of the employment agreement] or otherwise proprietary to Jostens."

[8] The trial court's Finding No. 28 explains that "[a] program or set of instructions with a beginning and a defined end, which can be invoked at its beginning and will always return to its end after performing a given function between those two points is called a `routine' or `sub-routine.' An example is a `square root routine' which will calculate the square root of a number when it is started and will produce the answer when it returns at its end."