CSS, Inc. v. Herrington et al, No. 2:2016cv01762 - Document 213 (S.D.W. Va. 2017)

Court Description: AMENDED MEMORANDUM OPINION AND ORDER denying the plaintiff's 65 MOTION for Preliminary Injunction. Signed by Judge Joseph R. Goodwin on 8/3/2017. (cc: counsel of record; any unrepresented parties) (taq)

Download PDF

CSS, Inc. v. Herrington et al Doc. 213 IN THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF WEST VIRGINIA CHARLESTON DIVISION CSS, INC., Plaintiff, v. CIVIL ACTION NO. 2:16-cv-01762 CHRISTOPHER HERRINGTON, et al., Defendants. AMENDED MEMORANDUM OPINION AND ORDER Pending before the court is Plaintiff’s Motion for Preliminary Injunction [ECF No. 65]. For the following reasons, the Motion is DENIED. I. Background The plaintiff, CSS, Inc. (“CSS”), has asked the court to preliminarily enjoin defendants, Christopher Herrington, Gene Yoho, and Compiled Technologies, LLC (collectively, “CT”), from “advertis[ing], licens[ing], offer[ing] to license, sell[ing] or otherwise mak[ing] available for use” CT’s land indexing and estate management software; “offer[ing] or . . . provid[ing] support services” for CSS’s software; or “mak[ing] use of any trade secrets, confidential information or proprietary information of CSS.” Pl.’s Mot. Prelim. Inj. 1 [ECF No. 65]. However, CSS is unable to show a likelihood of success on any of the claims presently before the court, that irreparable harm will occur absent the injunction, that the balance of the hardships weighs in its favor, nor that an injunction is in the public interest. Thus, a Dockets.Justia.com preliminary injunction is not warranted. II. Procedural History CSS filed its Complaint [ECF No. 1] on February 23, 2016. On September 16, 2016, CSS filed its First Amended Complaint. First Am. Compl. [ECF No. 59]. CSS brings actions against CT for copyright infringement, breach of contract, violation of the duty of loyalty, tortious interference with business relationships, and misappropriation of trade secrets under the West Virginia Uniform Trade Secrets Act (“WVUSTA”). In their Answer, CT filed a counterclaim against CSS for tortious interference. Answer 11 [ECF No. 61]. On November 23, 2016, CSS filed Plaintiff’s Motion for Preliminary Injunction wherein it asserted that it was likely to succeed on its copyright infringement, misappropriation of trade secrets claims, and breach of contract. Pl.’s Mot. Prelim. Inj. [ECF No. 65]. CT responded on December 13, 2016. Defs. Mem. Opp. Pls. Mot. Prelim. Inj. [ECF No. 68]. On December 16, 2016, CSS moved for an extension of time to reply to the defendants’ response. Unopposed Mot. for Extension of Time to File a Reply to Defs. Mem. Opp. Pls. Mot. Prelim. [ECF No. 72]. I granted this Motion [ECF No. 73], and CSS filed a Reply Brief [ECF No. 74] on December 29, 2016. On January 25, 2017, the court conducted a hearing on the Plaintiff’s Motion for Preliminary Injunction. Prelim. Inj. Mot. Hr’g [ECF No. 82]. Because of the volume and length of evidence put forth, the hearing could not be completed in one afternoon. This matter was continued to February 23, 2017, and it was ultimately completed on February 24, 2017. See Order Continuing Prelim. Inj. Hr’g 2 [ECF No. 84]; Mot. Hr’gs [ECF Nos. 89, 90]. On March 22, 2017, both parties filed their proposed findings of fact and conclusions of law. [ECF Nos. 106, 109, 110]. CSS’s Motion is ripe for my review. III. Findings of Fact I need only make findings of fact pertinent to determining whether a preliminary injunction should issue. I FIND that the facts of this case are as follows: 1. CSS is a company that provides software and related support services to local government entities, namely county clerks in West Virginia. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 28:1–9. 2. CSS has been in business since 1983. See Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 105:11. 3. Kofile acquired CSS in May 2014. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 39:18–19. 4. For over twenty years, CSS has provided various software solutions including: a land records indexing package, an estate management package as well as utility billing, sheriff’s tax collection, and other applications. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 28:1–4. 5. Defendant Christopher Herrington was first employed by CSS in October 1991. See Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 157:8–12, 190:14–15. 6. Christopher Herrington worked at CSS in the computer software, applications, and programming business related to county government information and document management systems. See Hr’g Tr., Feb. 23, 2017, C. Herrington 3 Direct, 157:8–12; see also Hr’g Tr., Jan. 25, 2017, Auburn Direct, 42:10–13. 7. During most of his employment with CSS, Christopher Herrington was a programmer, and his work included development of CSS’s software, bug fixes, and modifications. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 14:14–23, 16:22– 17:5. 8. Christopher Herrington also provided support for CSS software packages. Id. at 15:13–23. 9. Christopher Herrington’s duties included programming responsibilities and support responsibilities for both the CSS land indexing software and the CSS estate management software. Id. at 16:2–9. 10. On August 23, 2014, Christopher Herrington resigned as an employee of CSS. See Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 190:14–15. 11. On September 6, 2014, Christopher Herrington returned to CSS. See Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 203:22–24. 12. On September 12, 2014, Christopher Herrington was required to sign a confidentiality agreement. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 44:14– 45:2. 13. CSS had never required Christopher Herrington to sign a confidentiality agreement prior to September 12, 2014. See Hr’g Tr., Feb. 23, 2017, McCasker Redirect, 97:6–17; see also Hr’g Tr., Feb. 23, 2017, C. Herrington Cross, 239:3– 8. 14. Christopher Herrington was employed by CSS until June 5, 2015. See Hr’g Tr., 4 Feb. 23, 2017, C. Herrington Direct, 203:22–24. 15. On August 12, 2015, Gene Yoho formed CT. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 43:19–44:5. 16. Gene Yoho and Christopher Herrington operate CT. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 43:19–44:5. 17. CT is in the business of licensing custom computer software applications. CT offers one program for land records recording and indexing and one for an estate management application. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 44:3–5; see also Hr’g Tr., Feb. 23, 2017, C. Herrington Cross, 242:9–11. 18. CT also provides support services to county clerks’ offices related to these applications. See Hr’g Tr., Feb. 23, 2017, C. Herrington Cross, 242:9–11. 19. CT competes with CSS for West Virginia county clerk contracts. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 43:19–44:5. 20. CT’s code is not a literal1 copy of CSS’s code. See Mot. by Defs. for Leave to File Under Seal Ex. 1, at 26 [ECF No. 70-1] (“McCasker Expert Report”); see also Hr’g Tr., Feb. 23, 2017, McCasker Cross, 9:7–10. 21. The programming language RM/COBOL (“COBOL”) was used in the software of both CSS and CT. See McCasker Expert Report at 25; see also Mot. by Defs. for Leave to File Under Seal Ex. 2, at 3 [ECF No. 70-2] (“Zeidman Expert Report”). Literal copying is the wholesale copying of the software or its components, while non-literal copying is “copying that is paraphrased or loosely paraphrased rather than word for word.” Lotus Dev. Corp. v. Borland Int’l, Inc., 49 F.3d 807, 814 (1st Cir. 1995). 1 5 22. CSS uses COBOL version 11, and CT uses COBOL version 12. See McCasker Expert Report at 25. 23. CT’s COBOL data files are structured differently than CSS’s COBOL data files. See Hr’g Tr., Jan. 25, 2017, McCasker Direct, 115:14–15. 24. The parts of CSS’s software relevant to this case consist of the following architecture and technology: a client application written in the Microsoft Visual Basic 6 computer language (VB6), a VanGUI Network Communication Interface Builder (“VanGUI”), TCP (Transmission Control Protocol) Sockets, an application server with data files written in COBOL, a Relativity Server, an ODBC (Open Database Connectivity) Interface, and Crystal Reports OCX. See McCasker Expert Report at 7–9. 25. The parts of CT’s software relevant to this case consist of the following architecture and technology: a client application written in the C# computer language (pronounced “C-Sharp”), a Louis Network Communication Interface, TCP Sockets, an application server with data files written in COBOL, a Relativity Server, an ODBC Interface, and Crystal Reports. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 9–13. 26. CSS’s software includes third-party components such as VB6, the VanGUI interface, the COBOL programming language, the Relativity server, and Crystal Reports. See Zeidman Expert Report at 14–18. 27. The intellectual property rights to the third-party components—VB6, the VanGUI interface, the COBOL programming language, the Relativity Server, 6 and Crystal Reports—are not owned by any party to this case. See Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 148:21–149:12. 28. The source code was initially provided along with CSS’s software applications and installed on servers owned by the various West Virginia county clerks’ offices. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9, 35:12–18; Defs. Exs. 2, 14 from Prelim. Inj. Mot. Hr’g [ECF Nos. 96-2, 96-13]. 29. At some point after 2014, this practice was discontinued, and all of CSS’s source code was removed from all county servers. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9, 35:12–18. 30. Source code present on the county servers was available to anyone who had administrative rights2 authorized by the county. Id. at 31:3–20, 35:15–18. 31. Some of CSS’s contracts with county clerks’ offices did not have any language requiring the county clerks to protect CSS’s source code on the county servers. Id. 32. Some of CSS’s contracts with the county clerks’ offices did stipulate that the county must protect CSS’s source code. Id. at 32:12–22. 33. CSS’s corporate representative, Hubert “Bert” Auburn, does not know how many of CSS’s contracts required the clerks’ offices to protect CSS’s source code. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9; 35:15–18. 34. Bert Auburn does not know how many people had access to CSS’s source code. Id. at 35. 2 This refers to the ability of a person to access the files with an administrative login and password. 7 35. CSS founder, current employee, and father of defendant, Chris Herrington, Doug Herrington does not know how many people had access to CSS’s source code. See Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 105:12–13, 106:14–17, 120:13–16, 121:24–122:1. 36. Bert Auburn does not know whether CSS’s source code was passwordprotected. See Auburn Direct, 31:1–9; 35:15–18. 37. Bert Auburn is not aware of the source code being identified as confidential when on the county servers. See Mot. by Defs. for Leave to File Under Seal Ex. 4, at 60:13–18 [ECF No. 70-4] (“Auburn Dep.”). 38. Bert Auburn does not believe that CSS’s source code was encrypted when on county servers. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 35:1–9. 39. CSS’s source code was neither password protected nor encrypted on the county servers. See Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 120:21–23; Test. & Ex. Submission by CSS, Inc., Ex. 1 at 182:20–24 [ECF No. 105-1] (“Lowers Dep.”). 40. WVNet is a virtual private network that provides access to the county servers. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9, 35:15–18. 41. WVNet provides all companies with a contract with any county access to all West Virginia county servers. Id. 42. Anyone who can access WVNet can access any of the West Virginia county servers. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9, 35:15–18. 43. A competitor of CSS with a password for WVNet may have been able to access 8 CSS’s source code. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–20. 44. The password to access the CSS software application was “top gun.” See Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 135:3–11; see also Hr’g Tr., Feb. 23, 2017, Smith Direct, 175:20–22. 45. CSS did not enter into confidentiality agreements with the clerks’ offices that had the “top gun” password. See Hr’g Tr., Feb. 23, 2017, Smith Direct, 17–19. Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 136:20–23. 46. CSS would share the “top gun” password with county employees who attended training sessions. See Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 136:6–19; see also Hr’g Tr., Feb. 23, 2017, Smith Direct, 176:9–16. 47. CSS’s contracts with county clerks’ offices are subject to document requests pursuant to the state freedom of information act. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 38. 48. CSS’s pricing information for its county clients is public information. See Auburn Dep. 166:15–21. 49. CSS bases its pricing structure, in part, on a publically available tax base classification of counties that is published every two years. Id. at 167:14–15, 166:24–167:8, 179:5–6. 50. Christopher Herrington obtained CSS pricing information by email, face-toface conversations, and phone calls with county clerks. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 41:16–43:1. 51. CSS has some customers in Ohio. See Hr’g Tr. Auburn Direct, 28:9. 9 52. CT has no customers in Ohio. See Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 241:9–14. 53. CSS has lost contracts to CT. See Hr’g Tr., Jan. 25, 2017, 9:13–18. IV. Standard for Obtaining a Preliminary Injunction To secure a preliminary injunction, “plaintiffs must demonstrate that (1) they are likely to succeed on the merits; (2) they will likely suffer irreparable harm absent an injunction; (3) the balance of hardships weighs in their favor; and (4) the injunction is in the public interest.” League of Women Voters v. North Carolina, 769 F.3d 224, 236 (4th Cir. 2014) (citing Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008)). The Fourth Circuit reviews the grant or denial of a preliminary injunction for abuse of discretion. The Real Truth About Obama, Inc. v. Fed. Election Comm’n, 575 F.2d 342, 346–47 (4th Cir. 2009), vacated on other grounds, 559 U.S. 1089 (2010). “[A] a preliminary injunction is an extraordinary and drastic remedy, one that should not be granted unless the movant, by a clear showing, carries the burden of persuasion.” 11A Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2948 (3d. ed. 2017) (footnotes omitted). The Fourth Circuit permits district courts to rely on hearsay or other inadmissible evidence when deciding whether a preliminary injunction should be awarded. G.G. ex rel. Grimm v. Gloucester Cty. Sch. Bd., 822 F.3d 709, 725 (4th Cir. 2016), vacated on other grounds and remanded, 137 S. Ct. 1239 (2017). 10 V. Conclusions of Law A. Likelihood of Success on the Merits for the Copyright Claim In order to show copyright infringement, CSS must prove (1) it owned a valid copyright and (2) CT copied original elements of the copyright. Humphreys & Partners Architects, Ltd. P’ship v. Lessard Design, Inc., 790 F.3d 532, 537 (4th Cir. 2015) (citing Lyons P’ship v. Morris Costumes, Inc., 243 F.3d 789, 801 (4th Cir. 2001)). A valid copyright gives the owner the exclusive right to reproduce the copyrighted work and prepare derivative works. 17 U.S.C. § 106. Anyone who violates this exclusive right infringes the owner’s copyright. Id. at § 501(a). Copyright law protects “original works of authorship fixed in any tangible medium of expression.” 17 U.S.C. § 102(a). Computer programs are explicitly covered as “literary works.” 17 U.S.C. § 101; see Atari Games Corp. v. Nintendo of Am., Inc., 975 F.2d 832, 838 (Fed. Cir. 1992). Courts have also granted copyright protection over computer source code. See Whelan Assocs., Inc. v. Jaslow Dental Lab., Inc., 797 F.2d 1222, 1233 (3d Cir. 1986). Copyright protection extends to the “non-literal elements of computer programs that embody original expression”. Lotus Dev. Corp. v. Paperback Software Int’l, 740 F. Supp. 37, 76–77 (D. Mass. 1990). However, copyright analysis in the computer program context is often a “difficult task.” Oracle Am., Inc. v. Google Inc., 750 F.3d 1339, 1354 (Fed. Cir. 2014); see Lotus Dev., 49 F.3d at 820 (Boudin, J., concurring) (“Applying copyright law to computer programs is like assembling a jigsaw puzzle whose pieces do not quite fit.”); see also Comput. Assocs. Int’l, Inc. v. Altai, Inc., 982 F.2d 693, 696 (2d Cir. 1992) (“In 11 recent years, the growth of computer science has spawned a number of challenging legal questions, particularly in the field of copyright law. . . . As scientific knowledge advances, courts endeavor to keep pace, and sometimes—as in the area of computer technology—they are required to venture into less than familiar waters.”). As the Second Circuit remarked: To be frank, the exact contours of copyright protection for non-literal program structure are not completely clear. We trust that as future cases are decided, those limits will become better defined. Indeed, it may well be that the Copyright Act serves as a relatively weak barrier against public access to the theoretical interstices behind a program’s source and object codes. This results from the hybrid nature of a computer program, which, while it is literary expression, is also a highly functional, utilitarian component in the larger process of computing. Comput. Assocs. Int’l, 982 F.2d at 712. The scope of copyright protection, however, is limited in that it does not cover “any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such a work.” Id. at § 102(b); see Mazer v. Stein, 347 U.S. 201, 217 (1954) (“Unlike a patent, a copyright gives no exclusive right to the art disclosed; protection is given only to the expression of the idea—not the idea itself.”); see also Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 358 (1991); Avtec Sys., Inc. v. Peiffer, 21 F.3d 568, 572 (4th Cir. 1994) (“To reiterate, copyright protects not original ideas but their incarnation in a tangible means of expression.”). A work must be “original” to qualify for copyright protection. 17 U.S.C. § 102(a). This “originality requirement is not particularly stringent,” however. Feist Publ’ns, 499 U.S. at 358. In the copyright context, “original” means “that the work was 12 independently created by the author (as opposed to copied from other works), and that it possesses at least some minimal degree of creativity.” Id. at 345. The more creative a work is, the “thicker” the protection is for the work, and “less similarity is needed for a court to find the similarity sufficiently ‘substantial’ and thus infringing.” Lydia Pallas Loren, The Pope’s Copyright? Aligning Incentives with Reality by Using Creative Motivation to Shape Copyright Protection, 69 La. L. Rev. 1, 30 (2008). Conversely, highly functional, less creative works have “thin” protection, and more similarity is needed to show infringement. Id. 1. The Test For Copyright Infringement in the Fourth Circuit CSS is alleging non-literal copyright infringement, which covers “the structure, sequence, organization, user interface, screen displays, and menu structures.” Gen. Universal Sys., Inc. v. Lee, 379 F.3d 131, 142 (5th Cir. 2004). The Fourth Circuit has not yet explicitly endorsed a test for non-literal copyright infringement for computer source code. Many jurisdictions have now adopted the Second Circuit’s abstraction-filtration-comparison test (“AFC Test”) to deal with this particular type of work. 3 Robert W. Harris, A New Ball Game: Litigating Copyright Infringement for Nonliteral Software Elements, 76 J. Pat. & Trademark Off. Soc’y 157, 158 (1994) (“After a period of several years in which courts applied a rather bewildering variety of tests for substantial similarity of software products in copyright infringement cases, a consensus has recently emerged among Whether to adopt the AFC Test has previously been raised on appeal, but the Fourth Circuit did not decide this issue. Comprehensive Techn. Int’l, Inc. v. Software Artisans, Inc., 3 F.3d 730, 734–35 (4th Cir. 1993). 3 13 the circuits, on the proper approach for analyzing the substantial similarity of copyright-protectable elements of computer software, which approach was earlier recommended by the Nimmer copyright treatise.”). The Fifth, Ninth, Tenth, and Eleventh Circuits apply the AFC Test. See Gen. Universal Sys., 379 F.3d at 142 (“To assess a claim of software infringement, we have generally endorsed the ‘abstractionfiltration-comparison’ test first outlined by the Second Circuit in Altai and refined by the Tenth Circuit in Gates Rubber Co. v. Bando Chemical Industries, Ltd.”); Sega Enters. Ltd. v. Accolade, Inc., 977 F.2d 1510, 1525 (9th Cir.1992) (“In our view, in light of the essentially utilitarian nature of computer programs, the Second Circuit’s approach is an appropriate one.”); Gates Rubber Co. v. Bandon Chem. Indus, Ltd., 9 F.3d 823, 833 (10th Cir. 1993); MiTek Holdings, Inc. v. Arce Eng’g Co., 89 F.3d 1548, 1559 (11th Cir. 1996) (“Our circuit, in applying the Altai test, employs the substantial similarity standard in comparing what remains after the abstraction and filtration steps . . . .”). The Sixth Circuit uses a similar test, and the First Circuit addressed the AFC Test in its case of first impression and viewed it favorably.4 Kevin J. Hickey, Reframing Similarity Analysis in Copyright, 93 Wash. U.L. Rev. 681, 694 (2016); see, e.g., Lexmark Int’l, Inc. v. Static Control Components, Inc., 387 F.3d 522, 534 (6th Cir. 2004); Lotus Dev., 49 F.3d at 815. The AFC Test “has the major advantage of being entirely consistent with the infringement calculus used for infringement issues concerning other types of works. However, the First Circuit ultimately did not rely on the AFC Test because Lotus involved literal infringement. Lotus Dev., 49 F.3d at 814–15. 4 14 In effect, it is an adaptation of the traditional infringement test to the medium of computer programs.” Howard B. Abrams, Copying of Protected Expression— Technology, Functionality and the Ordinary Observer Test—Commentary, 2 The Law of Copyright § 14:33 (2016). The AFC Test “arose out of the difficulties of assessing similarity in computer software cases. At its core, the AFC Test is simply a formalized system to ensure that elements that are not protected by copyright are excluded when comparing two works.” Hickey, supra, at 694. In essence, the AFC Test works as follows: In ascertaining substantial similarity [between two or more programs] under this approach, a court would first break down the allegedly infringed program into its constituent structural parts. Then, by examining each of these parts for such things as incorporated ideas, expression that is necessarily incidental to those ideas, and elements that are taken from the public domain, a court would then be able to sift out all non-protectable material. Left with a kernel, or possibly kernels, of creative expression after following this process of elimination, the court’s last step would be to compare this material with the structure of an allegedly infringing program. The result of this comparison will determine whether the protectable elements of the programs at issue are substantially similar so as to warrant a finding of infringement. Comput. Assocs. Int’l, 982 F.2d at 706. I will therefore employ the AFC Test to determine substantial similarity for alleged non-literal copyright infringement of source code. 2. Relevant Background Information Given the complexity of this type of copyright infringement analysis, some background information is helpful. In this case, both CSS and CT’s programs have client/server architecture. Client/server describes “a program relationship in which the client requests a service or resource from the server.” Zeidman Expert Report, Ex. 15 J at 1. Both parties’ programs are client/server applications, meaning that there is a “front end” or “client side” and a “back end” or “server” side. Id.; McCasker Expert Report at 7, 9. The three coding languages used in this case are COBOL, VB6, and C#. The server sides of both CSS and CT are written in COBOL. The client side of CSS is written in VB6, and the client side of CT is written in C#. COBOL is “exclusively owned, developed and maintained by Micro Focus.” Defs. Mem. Opp. Pls. Mot. Prelim. Inj., Ex. 9, at 2 [ECF 68-9] (“Letter from Courtney Wood, Associate General Counsel for Micro Focus”). This case also involves several third-party applications: Relativity, VanGUI, Louis, and Crystal Reports. Relativity is a data-access middleware product, specifically it is an ODBC-accessible relational database management system. Defs. Mem. Opp. Pls. Mot. Prelim. Inj., Ex. 9, at 4–5 [ECF 68-9]. It improves access to COBOL data. Relativity is a COBOL-specific database engine/driver that is “exclusively owned, developed and maintained by Micro Focus.” 5 Letter from Courtney Wood, Associate General Counsel for Micro Focus 2. VanGUI is a communication middleware that allows application programs written in COBOL to be combined with graphical user interface (GUI) front-ends written in programming languages that support VB6. Defs. Mem. Opp. Pls. Mot. Prelim. Inj., Ex. 9 at 6–7 [ECF 68-9]. VanGUI is “a way to modernize COBOL applications.” Id. at 1; Zeidman Liant acquired the rights to use and market Louis with its RM/COBOL and other COBOL products prior to its acquisition by Micro Focus in 2008. 5 16 Expert Report at 18. Louis is an interface builder that works in conjunction with COBOL platforms and products and GUIs. Louis is “exclusively owned, developed and maintained by Micro Focus.” Letter from Courtney Wood, Associate General Counsel for Micro Focus 2. Crystal Reports is a business intelligence application that helps generate reports. Zeidman Expert Report, Ex. H at 2. 3. CSS has Valid Copyrights A certificate of copyright registration is prima facie evidence of ownership of a valid copyright. 17 U.S.C. § 410(c); Serv. & Training, Inc. v. Data Gen. Corp., 963 F.2d 680, 688 (4th Cir. 1992). CSS is the registered owner of three copyrights in computer software code: Document Indexing and Imaging for Counties 2015 (TX 8230-758), Estate Management 2015 (TX 8-230-809), and Web Inquiry 2013 (TX 8-230816). CT did not dispute the validity of the copyrights in its Response to CSS’s motion. The defendants did not introduce evidence calling the validity of the copyright into question.6 If CT had wished to dispute validity at the hearing, it would have had the “burden of overcoming the presumption [of validity] arising out of the granting of the copyright by the Copyright Office.” M. Kramer Mfg. Co. v. Andrews, 783 F.2d 421, 434 (4th Cir. 1986). Thus, I FIND that CSS is likely to show that it owns valid copyrights. 4. The AFC Test The next stage of the analysis is more complex. In order to demonstrate Through cross examination, defendants have implied that the dates of creation on the estate and indexing software copyright registrations are incorrect. Hr’g, Tr., C. Herrington Direct, 233: 3–234:2. However, they did not present sufficient evidence to rebut the presumption of validity. 6 17 copyright infringement, CSS must produce either direct evidence of infringement, or in the alternative, that (1) CT had access to the copyrighted work, and (2) CT’s software is substantially similar to CSS’s copyrighted material. Comput. Assocs. Int’l, 982 F.2d at 701. CSS argues the latter, which means I must apply the AFC Test for substantial similarity. a. Step 1: Abstraction The first step described in Altai is abstraction. Abstraction is a very factspecific process. The overarching goal of abstraction is to break out a program into its numerous ideas and expressions. Id. at 706–07. Each level of the AFC Test involves “dissective analysis.” Pamela Samuelson, A Fresh Look for Nonliteral Copyright Infringement, 107 Nw. U. L. Rev. 1821, 1838 (2013). The purpose of this step is to distinguish the copyrightable expressive aspects of software from its underlying ideas, which are not protectable by copyright. Id. at 707. Analytic dissection is relevant not only to the copying element of a copyright infringement claim, but also to the claim’s ownership element. One aspect of the ownership element is the copyrightability of the subject matter and, more particularly, the scope of whatever copyright lies therein. To the extent a plaintiff’s work is unprotected or unprotectable under copyright, the scope of the copyright must be limited. Brown Bag Software v. Symantec Corp., 960 F.2d 1465, 1476 (9th Cir. 1992) (citations omitted). I rely entirely on the report and testimony of CSS’s expert, Andrew McCasker, and the report of CT’s expert, Dr. Robert Zeidman to perform the abstraction process. b. Step 2: Filtration The second step in the AFC Test is filtration. After the experts identify the 18 abstracted levels of the software in this case, the court must next examine the structural components and determine if they are copyrightable. Comput. Assocs. Int’l, 982 F.2d at 707. The purpose of filtration is to define the scope of the plaintiff’s available copyright. Id. Courts have been encouraged to be more explicit about addressing which elements in protected works are not protectable by copyright. See Samuelson, supra, at 1839. The Altai case names three specific categories of elements that cannot be protected: (1) elements dictated by efficiency, which are barred by the merger doctrine, (2) elements dictated by external factors, similar to scènes à faire in traditional copyright analysis, and (3) elements from the public domain. Comput. Assocs. Int’l, 982 F.2d at 707–11. When there is only one way to express a particular idea, the idea and its expression are “merged” and copyright will not protect that expression. Id. Coders are driven to write the leanest, most efficient code in order to make a program run as fast as possible.7 Efficiency is an industry-wide goal. Since, as we have already noted, there may be only a limited number of efficient implementations for any given program task, it is quite possible that multiple programmers, working independently, will design the identical method employed in the allegedly infringed work. Of course, if this is the case, there is no copyright infringement. “Software is designed and written with the goals of limiting the amount of computer memory required, accomplishing the task quickly, shortening the design and coding phases, and establishing an easy to understand user interface.” Crowe, supra, at 212. Coders strive for “program efficiency” in their applications: the fastest processing speed, lowest memory utilization, and “the quality and speed of the input/output functions.” See Peter S. Menell, An Analysis of the Scope of Copyright Protection for Application Programs, 41 Stan. L. Rev. 1045, 1051 (1989). 7 19 Comput. Assocs. Int’l, 982 F.2d at 708 (citations omitted) (emphasis added). “[E]vidence of similarly efficient structure is not particularly probative of copying[;] it should be disregarded in the overall substantial similarity analysis.” Id. at 709. The scènes à faire doctrine refers to standardized features that are not copyrightable. This doctrine also excludes from copyright protection work serving functional purposes or work that is dictated by external factors such as particular business practices. Comp. Mgmt. Assistance Co. v. Robert F. DeCastro, Inc., 220 F.3d 396, 401 (5th Cir. 2000). Courts are to consider how a programmer’s freedom of choice might be impeded by external considerations such as the mechanical specifications of the computer on which the code will run, compatibility requirements with other programs, the computer manufacturer’s design standards, industry demands, and widely accepted programming practices. Comput. Assocs. Int’l, 982 F.2d at 710. Elements from the public domain are any elements of a computer program that may have entered the public domain. Id. at 710–11. They cannot be appropriated by an author even if they are included in a copyrighted work. Id. c. Step 3: Comparison The final step is to compare the expressive elements of the plaintiff’s software that remain with the elements of the defendant’s program alleged to infringe. Finding infringement is warranted if there is substantial similarity in the programs’ expressions and those similarities were the result of copying. Id. 5. Alleged Copyright Infringement Mr. McCasker asserts that CT has infringed the “overall structure, 20 architecture, and flow of CSS[’s] programs.” McCasker Expert Report at 2. He created schematics of each company’s programs, 8 and stated that “[b]y visually scanning these figures together, the degree to which the architecture components of the two applications are the same is readily apparent.” McCasker Expert Report at 13. CSS asserts that “[n]onliteral elements of a computer program may receive copyright protection even if they are individually unprotectable, if they are compiled in a unique or creative way.” SecureInfo Corp. v. Telos Corp., 387 F. Supp. 2d 593, 612 (E.D. Va. 2005). Through Mr. McCasker’s expert report and the evidence given at the hearing, CSS also alleges, inter alia, the following specific characteristics of CT’s software constitute non-literal infringement: the decision to use temporary files for the reporting function; the use of middle level of modularity for the business functions9; the use of one-to-one level of modularity for the reporting functions; the use of the program name to call services from the back end with the functions to be performed, rather than the “superior” approach of passing only the function to be carried out and letting the server determine the program to run; the “direct call” transaction flow in the business portion; the fact both companies use Crystal Reports, Relativity, and VanGUI or Louis; and the use of a letters-only name search algorithm. McCasker Expert Report at 3–10; Hr’g Tr., McCasker Direct, Jan. 25, 2017, 107:21–25; Hr’g Tr., Dr. Zeidman notes that it hasn’t been shown that Mr. McCasker’s representation of the architecture “actually represents the functionality of the code.” Zeidman Dep. 75:10–16 [ECF No. 105-2]. 8 The business process in both programs use middle level of modularity, meaning there are multiple actions or grouped actions carried out by a single COBOL program on the server side. Hr’g Tr., Feb. 23, 2017, at 81:18–25, 88:15–21. 9 21 Feb. 23, 2017, McCasker Redirect, 81:18–25, 87:5–8, McCasker Cross, 133:21–25. I will not address every single allegation in detail, but instead focus on the overall problems that CSS has in meeting its burden. Mr. McCasker’s testimony was genuine, and I found him to be sincere and knowledgeable. However, I found him unable to show substantial similarity using the AFC Test. a. The Choice of COBOL CSS has indicated that the choice to use COBOL supports the argument that CT has infringed its software programs, despite the fact that COBOL is not the intellectual property of CSS. CSS alleged through its expert that the choice of COBOL was a part of the infringement analysis: THE COURT: Choosing a language the same as somebody else chose the same language, you’re not saying that that’s part of the problem, are you? MR. MCCASKER: I’m saying that the choice of the language drives architecture decisions, and there are other choices available to them that would have resulted in a different architecture. Hr’g Tr., Jan. 25, 2017, McCasker Direct, 124:8–14. CSS spent a great deal of time throughout the three days of the hearing, in moving papers, and in depositions maligning COBOL as a very old, outdated language that no one10 uses any more. See, e.g., H’rg Tr., Jan. 25, 2017, Auburn Direct, 41:4 (“COBOL is a very old language.”); McCasker Expert Report at 4 (“COBOL is an outdated language.”); McCasker Expert Report at 25 (“Using RM/COBOL to build a new application is almost unheard of in the last 20 years.”); 10 Additionally, because CSS’s own software is written in COBOL, this argument is not compelling. 22 Hr’g Tr., Jan. 25, 2017, McCasker Direct, 98:4–5 (“Are you old enough—not to be insulted—to have programmed in COBOL?”); Id. at 108:13–15 (“[T]here are other languages that are generally considered better and more modern. That’s one of the reasons why COBOL programming is in decline.”). Because the choice of coding language is not protected by CSS’s copyrights, this was not time well spent. Additionally, CSS’s COBOL database and CT’s COBOL database are not the same. Hr’g Tr., Jan. 25, 2017, McCasker Direct, 111:16–17. Mr. McCasker noted during the hearing that structure of the COBOL files is “similar but different.” Id. at 111:19–25. I FIND that the choice of COBOL is not evidence of substantial similarity. b. Structure, Architecture, and Design CSS’s allegation regarding the similarity of the software applications’ structure and design suffers from many problems. The first is that CSS is unable to perform the abstraction step of the AFC Test. CSS does not “distinguish the copyrightable expressive aspects of software from its underlying ideas, which are not protectable by copyright.” Comput. Assocs. Int’l, 982 F.2d at 707. Copyright protection is limited “to the expression of an idea and not the underlying idea itself.” Menell, supra, at 1104 n.7. “Because computer programs are utilitarian works, the process-expression dichotomy is an important limit on the scope of their protection.” Steven R. Englund, Idea, Process, or Protected Expression?: Determining the Scope of Copyright Protection of the Structure of Computer Programs, 88 Mich. L. Rev. 866, 901–02 (1990). Though software “design” as an aesthetic can be protectable, the way 23 CSS refers to “design” is how the programs function. Mr. McCasker describes the architecture and specific processes he believes are copied in terms of their functionality rather their expression. Zeidman Dep. 73:14– 18; Zeidman Expert Report at 12. In his expert report, Mr. McCasker defines a program’s structure as follows: For the purpose of this expert report, the definition I am using for software architecture comes from Gates Rubber v. Bando Chemical: The program’s architecture or structure is a description of how the program operates in terms of its various functions, which are performed by discrete modules, and how each of these modules interact with each other. McCasker Expert Report at 6. Thus, he found substantial similarity in terms of how the programs acted, and not how the programs were expressed. McCasker Expert Report at 15–16. For example, when discussing the Indexing software, he remarked, “[t]he COBOL program extracts the action from the request array, as well as the parameters. [] The COBOL program reads, writes, updates, and/or deletes COBOL permanent data files records.” Id. The same is true of his descriptions of the transaction flow of the parties’ other applications. CSS fails to perform the abstraction part of the AFC Test correctly. The second problem is that Mr. McCasker’s report did not identify specific files or code at issue. Zeidman Expert Report at 8. In any case, he did not “do so with enough specificity to be understood.” See, e.g., id. at 11. (“Mr. McCasker does not state whether Inquiry is a functional executable file, one or more source code files, or a set of functions in the source code.”). Though this is a case regarding non-literal infringement, CSS still needed to show specifically which protectable, expressive 24 elements of its software CT had infringed. Thus, I FIND that CSS is not likely to succeed on the merits of its copyright infringement claim with respect to the structure, design, architecture, or design of the parties’ software. c. Third-Party Components and Transaction Flow CSS’s allegations regarding the third-party applications and transaction flow are not viable because third-party components must be filtered out in the second step of the AFC Test. Comput. Assocs. Int’l, 982 F.2d at 707–11. Thus, even though CT employed identical and similar applications in its software—such as Relativity or Crystal Reports—this is not infringement, because CSS does not own the intellectual property for those applications. The choice of COBOL played a critical role in defining which third-party applications CT needed to use in order to run its programs effectively. As Mr. McCasker noted, “the choice of the language drives architecture decisions.” Hr’g Tr., Jan. 25, 2017, McCasker Direct, 124:11–14. He testified that “the choice of the language dictated . . . the choice of components in between that in turn drive design decisions that—where the design ends up being the same.” Id. at 125:2:5. Because the choice of COBOL necessitated the use of the specific third-party applications CT employed, they should be filtered out because of the scènes à faire doctrine. Comput. Assocs. Int’l, 982 F.2d at 710. As Mr. McCasker noted, “the ‘transaction flow’ is dictated for the most part by the use of the components, their requirements, and how they interact with each other 25 from end-to-end through the applications.” McCasker Expert Report at 6. For example, “there are some elements of the design that the choice of COBOL is sort of— makes a natural choice, like a VanGUI interface.” Hr’g Tr., Feb. 23, 2017, McCasker Direct, 29:4–6. Doug Herrington testified, however, that he was not aware of another method to connect a VB6 front end to the COBOL back end of a software program other than with the VanGUI interface. Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 148:16–20. Also, there is appears to be no method to connect a C# front end of a software program to the COBOL back end of a software program other than with the Louis interface, used by CT in the same place as VanGUI.11 Additionally, Relativity requires an ODBC interface12 and COBOL data files. Zeidman Expert Report at 15. A TCP socket is a computer industry “standard protocol for exchanging data between computer applications over a network.” See Zeidman Expert Report at 16–17. These elements are not protectable by copyright, and they are to be filtered out in the second step of the AFC Test. I FIND that neither the choice of third-party components, nor the arrangement of those third-party components, nor the similar transaction flow of the parties’ software is evidence of substantial similarity. d. Specific Allegations of Non-Literal Infringement McCasker states that CT had to “custom code” their own version to use the VanGui technology. McCasker Expert Report at 25; Hr’g Tr., Jan. 25, 2017, McCasker Direct, 123:15–22. However, at the preliminary hearing it was established that the defendants utilized Louis, an application created by Liant, which is now owned by Micro Focus. Letter from Courtney Wood, Associate General Counsel for Micro Focus 2–3. 11 An ODBC Interface is a computer industry standard tool commonly used by computer programmers to develop computer applications. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 12–13. 12 26 CSS fails to show that CT’s software packages are substantially similar to the copyrightable elements of its software. For a defendant to be liable for copyright infringement, “the court must find that the defendant copied protectable elements of the plaintiff’s program and that those protectable elements comprise a substantial part of the plaintiff’s program when it is considered as a whole.” Gates Rubber, 9 F.3d at 833 (emphasis added). “When similar works resemble each other only in those unprotected aspects, then defendant prevails.” Melvin B. Nimmer & David Nimmer, 4 Nimmer on Copyright § 13.03 (B)(2) (internal citations omitted). “Infringement is shown by a substantial similarity of protectable expression, not just an overall similarity between works. Thus before evaluating substantial similarity, it is necessary to eliminate from consideration those elements of a [computer] program that are not protected by copyright.” Id. at § 13.03(F). With regard to many of the specifically alleged infringement specifics, most of what CSS has alleged are merely the most efficient or effective ways to implement a task. The merger doctrine requires evidence of the “similarly efficient structure[s]” to be disregarded and filtered out in the second step of the AFC Test because the function and expression are “merged.” Comput. Assocs. Int’l, 982 F.2d at 708. Thus, despite CT’s software having similar or even identical choices to that by CSS, these expressions are not protectable.13 For example, as Mr. McCasker asserted: defendants’ program copies CSS’s “Merely asking the question, ‘Does another way exist to perform a particular function?,’ does not resolve the idea/expression puzzle. Some ways are better than others. And with respect to computer programs, ‘better’ typically means more efficient—e.g., faster execution speed, less memory utilization, compatibility with data storage methods.” Menell, supra, at 1085. 13 27 letters-only name search algorithm that allows for more accurate searches in county records. Hr’g Tr., Jan. 25, 2017, McCasker Direct, 135:12–136:12 (“[I]t increases the likelihood that you’ll get a match and it avoids problems . . . . It’s a more accurate search.”). The use of temporary COBOL data files was also described to be a common and more efficient way to access data. Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 35:10–37:2 (“COBOL has created temporary files for as long as it’s been in existence. . . . So it’s more efficient.”). Thus, these elements are barred by the merger doctrine and are not copyrightable. With regard to the “unusual” combination of password and login, Mr. McCasker testified that “Windows XP actually has the same kind of user interface.” Id. 138: 3–4. Dr. Zeidman noted in his expert report that, “[t]his is evidence that the technique is well-known in the industry . . . .” Zeidman Expert Report at 33. Again, when a particular way for software to run is the most efficient, which can be shown by common usage—the expression and idea merge—and thus, that part is not protectable by copyright. See Comput. Assocs. Int’l, 982 F.2d at 707–11 For the additional allegations, CSS has failed to show similarity at all between its source code and that of CT. There was only one time in the course of this hearing that lines of source code were discussed in any detail, and this occurred on the third day of the hearing when CT introduced Exhibit 11 and Exhibit 12 through Christopher Herrington. Defs. Exs. 11, 12 from Prelim. Inj. Mot. Hr’g [ECF Nos. 9611, 96-12]. There it was established that the parties made very different 28 programming choices in writing their code for the process of invoking a program on the server side. Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 16:5–31–22. When the client application “request[s] a COBOL program to be executed, it does so by populating an array.” Hr’g Tr., Feb. 23, 2017, McCasker Cross 14:7–22. CSS alleges that “[t]he way in which that array is populated is the same in both programs.” Id. Mr. McCasker stated that both parties “embed the name of the COBOL program to be launched.” Id. at 18:1–17. However, Christopher Herrington, and the exhibits themselves, showed that in fact the CSS program makes a single array composed of 100 positions (lines 0 to 99), while the CT program creates two separate arrays of five lines each (lines 0 to 4) and packages them into one transaction. Compare Ex. 11, at 1 with Ex. 11, at 2; see also Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 23:18–25. Mr. McCasker also testified that both programs populate the name of the COBOL program into the array in the first position. Hr’g Tr., Feb. 23, 2017, McCasker Cross, 28:4–8. But the code and later testimony revealed that CSS’s code places the program name in the 100th position. See Ex. 11, at 1 (“VGArrayPutString vgvaction, 99, ‘dimnt.cob’”). CT has its COBOL program name in the null position, in line 0. See Ex. 11, at 2 (a[0] = ‘cbo.cob’;”). The way each program is expressed is thus very different. Exhibit 12 illustrates another example of the parties’ different coding choices for invoking a COBOL program on the server side. In this section of the source code, CSS again has chosen to create two arrays. See Ex. 12, at 1. The first array is sent by 29 “VGMakeArray vgvacation”, and the second is sent by “VGMakeArray vgvRecord.” Id. CSS places the COBOL name in position 99 out of 100 positions. Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 31:6–9. The CSS code passes many parameters (such as “From Date,” “Format Date,” “Thru Date”), while CT’s code does not pass any parameters. Id. at 32:10–24; Ex. 12, at 1–2. What this means is that CSS has actually hard-coded its parameters directly into the code, while CT’s parameters are “read from a file and retrieved.” Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 33:2–23. This is a completely different design decision. Mr. McCasker testified that these are exceptions—that he missed in his analysis of the source code—to what he normally observed in CT’s and CSS’s arrays with “program name being first and the action is second.” Hr’g Tr., Feb. 24, 2017, McCasker Direct, 78:15–20. Nevertheless, Mr. McCasker was unable to determine how often this was the case, and CSS did not present to any particular lines of code to show this was indeed an exception. I FIND that CSS is not likely to succeed on any of its copyright infringement claims. B. Likelihood of Success on the Merits for Misappropriation of Trade Secrets CSS alleges several trade secrets that CT is to have purportedly used in creating and selling its software. The claims can be broken down into two categories: alleged trade secrets related to CSS’s software and alleged trade secrets related to CSS’s business practices. Specifically, the plaintiff asserts that CT misappropriated its source code, software upgrades, method of pricing, business strategy, marketing, 30 geographic expansion, customer preferences, and client list. Pl. Proposed Findings of Fact & Conclusions of Law 14 [ECF No. 106]. 1. The Six-Factor Test for Theft of Trade Secrets West Virginia has adopted the Uniform Trade Secrets Act codified as the West Virginia Uniform Trade Secrets Act (“WVUSTA”). W. Va. Code § 47–22–9. In determining whether a violation of the WVUSTA has occurred, the court must first determine whether there was a trade secret to be protected, and then—if a trade secret exists—it must determine whether it has been misappropriated. W.Va. Code Ann. § 47–22–1(d); McGough v. Nalco Co., 496 F. Supp. 2d 729, 739 (N.D. W. Va. 2007). The plaintiff bears the burden of producing evidence that what it is alleging meets the definition of a trade secret. Trandes Corp. v. Guy F. Atkinson Co., 996 F.2d 655, 661–62 (4th Cir. 1993); IVS Hydro, Inc. v. Robinson, 93 F. App’x 521, 526–27 (4th Cir. 2004); McGough, 496 F. Supp. 2d at 739. Under this act, a trade secret is defined as “information” that (1) derives “independent economic value” from “not being generally known” or “readily ascertainable” by others who can obtain economic value from its use or disclosure, and (2) is the subject of “efforts that are reasonable under the circumstances to maintain” the secrecy of the information. W. Va. Code §§ 47–22–1(d)(1)–(2); IVS Hydro, 93 F. App’x at 526. In State ex rel. Johnson v. Tsapis, 419 S.E.2d 1 (W. Va. 1992), the Supreme Court of Appeals of West Virginia adopted the six-factor test found in § 757 of the Restatement of Torts to determine whether there was good cause to issue a protective order to prevent the disclosure of the defendant’s trade secrets. 31 The factors set forth in Tsapis are as follows: (1) the extent to which the information is known outside of the defendant’s business; (2) the extent to which it is known by employees and others involved in the defendant’s business; (3) the extent of the measures taken by the defendant to guard the secrecy of the information; (4) the value of the information to the defendant and competitors; (5) the amount of effort or money expended by the defendant in developing the information; and (6) the ease or difficulty with which the information could be properly acquired or duplicated by others. IVS Hydro, 93 F. App’x at 526–27; McGough, 496 F. Supp. 2d at 739. “[T]he absence of evidence as to any one factor does not necessarily preclude a finding that a trade secret exists. The factors should be viewed as instructive guidelines, weighed together in making that determination.” McGough, 496 F. Supp. 2d at 740 (internal citations omitted). The hallmark of a trade secret is not its novelty but its secrecy. Avtec Sys., 21 F.3d at 575. Secrecy is a question of fact. Trandes, 996 F.2d at 664. “Absolute secrecy is not essential,” and limited disclosure does not destroy secrecy. See id. (“A trade secret owner, however, does not abandon his secret by a limited public publication for a restricted purpose.”). The Restatement (First) of Torts advises that to be a trade secret, “a substantial element of secrecy must exist, so that, except by the use of improper means, there would be difficulty in acquiring the information.” Restatement (First) of Torts § 757 cmt. b (Am. Law Inst. 2017); Avtec Sys., 21 F.3d at 575. 32 A trade secret is misappropriated when it is improperly acquired or disclosed by someone who knew, or had reason to know, of a duty to maintain confidentiality. Id. 574–5; W. Va. Code Ann. § 47-22-1(b). 2. Source Code and Software-Related Claims Initially, the source code files for CSS’s software were loaded onto each client’s servers so in the event CSS went out of business, the client’s work would not be disrupted. Hr’g Tr., Auburn Direct, Jan. 25, 2017, 35:12–25. CSS provided the source code as a part of its software package order. See, e.g., Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:1–9 and 35:15–18, Defs. Exs. 1, 2, 14 from Prelim. Inj. Mot. Hr’g [ECF Nos. 96-1, 96-2, 96-13]. Bert Auburn stated that this “bad business practice” has not been continued, and all of CSS’s source code has now been removed from all county servers. Auburn Dep. 58:21–60:19. However, this removal could not have occurred until Bert Auburn came on board, which would be at some point after CSS was sold to Kofile in May 2014. Hr’g Tr., Jan. 25, 2017, Auburn Direct, 39:18–19. Both Doug Herrington and Bert Auburn were unable to state with certainty who had access to CSS’s unprotected source code from 1983 until 2014. CSS. Id. at 60:13–18; Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 120:13–16, 121:24–122. What I find especially injurious to CSS’s claim is that West Virginia county servers are all connected through WVNet, “the virtual private network that allowed vendors to access clients [and] state-run computers.” Defs.’ Mem. Opp’n Pl.’s Mot. Prelim. Inj., Ex. 3, at 267:7–15. Any and all county vendors can access the servers, including competitors of CSS. Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 119:9– 33 11. CSS left decisions regarding who could access its source code on the servers entirely up to the county clients themselves. The county clients—not CSS—were the ones who determined who had “admin rights”14 in order to access the files: MR. HERRING: And it was y’all’s understanding that the counties were taking appropriate measures to safeguard that system? BERT AUBURN: Yes, it was our understanding, our belief, and that’s what we had discussed with clients. Hr’g Tr., Jan. 25, 2017, Auburn Direct, 31:21–24. However, CSS did not uniformly mandate any type of confidentiality requirements when it provided the county clients its software nor did CSS otherwise require that the source code be kept confidential. See Hr’g Tr., Jan. 25, 2017, Auburn Direct, 32:12–24. Bert Auburn did not know if any of the county contracts had language that required protection of CSS source code. Id.; Hr’g Tr., Jan. 25, 2017, Auburn Cross, 75:8–14. When asked what limits, if any, CSS itself put on its software to source code to limit access, Doug Herrington replied “[n]othing; nothing particularly on the source code.” Id. at 120:17–20. CSS provided source code to clients in a way that “[the files] were pretty much editable in Notepad or however else you needed to view them.” Lowers Dep. 182:20–24; Auburn Dep. 58:21–60:19. Bert Auburn did not know at his deposition whether the server was in any way “locked,” whether the source code was password protected, nor whether the source code was identified as confidential. Id. at 60:13–18. While Doug Herrington testified that it was neither password protected 14 This is a reference to “administrative rights.” 34 nor encrypted. CSS. Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 120:21–23. Additionally, to log into CSS’s applications, Doug Herrington agreed that CSS maintained the same password, “top gun,” to log into its applications for likely more than twenty-five years. Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 135:3–136:14. The password was told to anyone who attended a training, and CSS did not require the password to be kept confidential. See Hr’g Tr., Feb. 23, 2017, D. Herrington Cross, 136:6–19; see also Hr’g Tr., Feb. 23, 2017, Smith Direct, 176:9–16. Thus, when I consider the “ease or difficulty with which the information could be properly acquired or duplicated by others,” CSS fails to pass muster. IVS Hydro, 93 F. App’x at 526–27; McGough, 496 F. Supp. 2d at 739. CSS argues that the defendants were unable to prove that anyone had in fact accessed the code, but this misstates what is required. CSS’s vigorous litigation of this case is not able to undo decades of neglecting to protect its source code. When considering the “extent to which the alleged trade secret is known by employees and others involved in [CSS’s] business,” CSS also did not make reasonable efforts to protect its confidential material until recently. IVS Hydro, 93 F. App’x at 526–27; McGough, 496 F. Supp. 2d at 739. Christopher Herrington,15 an employee since 1991, was never made to sign a confidentiality agreement until 2014. Pl. Mot. Prelim. Inj. Ex. 2 [ECF 65-2]; Hr’g Tr., Jan. 25, 2017, Auburn Direct, 45:10– 21. 15 Although he did concede at the hearing that he know the source code was confidential. Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 159:4–7. 35 After considering the six Tsapis factors, I FIND that CSS has not made a prima facie case of misappropriation with respect to its source code under the WVUSTA. Trandes, 996 F.2d at 661–62; IVS Hydro, 93 F. App’x at 526–27; McGough, 496 F. Supp. 2d at 739. CSS has failed to meet its burden of proving a trade secret exists. Thus, I FIND CSS’s source code does not meet the definition of a trade secret, and CSS is not likely to succeed on the merits of its WVUSTA claim with respect to its source code. 3. Business Practice-Related Claims a. CSS’s Client List CSS’s client list cannot be considered a trade secret because the names of CSS’s clients are publically available information. CSS’s contracts are with public entities subject to document requests pursuant to the West Virginia Freedom of Information Act. See W. Va. Code § 29B–1–2-(5). The names of clients are available to anyone who requests the information. In Avtec Sys., Inc. v. Peiffer, the Fourth Circuit was not able to find a case that supported the idea that an alleged owner of a trade secret “could maintain the secrecy of material that is subject . . . to publication at the will of another.” Avtec Sys., 21 F.3d at 575. The West Virginia Supreme Court of Appeals has also found that customer lists do not “constitute[] a protectable employer interest.” Memo. Op. and Order at 12, IVS Hydro, Inc. v. Robinson, No. 2:01-cv-1296 (S.D. W. Va. May 29, 2003) (citing Voorhees v. Guyan Machinery Co., 446 S.E. 2d 677 (W. Va. 1994)). I found previously in McGough that “customer lists [would not] be a trade secret” when the accounts are 36 “pretty much common knowledge” in the industry. McGough, 496 F. Supp. 2d at 740. The evidence also suggests that CT directly solicited all counties in West Virginia, not just those that were clients of CSS. Defs. Ex. 14 from Prelim. Inj. Mot. Hr’g [ECF No. 96-14]; Hr’g Tr., Feb. 23, 2017, C. Herrington Direct, 228:1–2. Thus, I FIND that CSS is not likely to succeed on the merits of its WVUSTA claim with respect to customer names. a. CSS’s Pricing Methodology CSS’s pricing methodology does not meet the definition of trade secret. The Fourth Circuit has held that pricing may not be a trade secret when the “approach to pricing [is] not unique and [the company] did not expend a great deal of money developing its pricing techniques.” IVS Hydro, 93 F. App’x at 528. Bert Auburn testified that their pricing methodology is based on the tax base in the county, which is determined by publications that are “completely public information.” Auburn Dep. 166:24–165:8; Hr’g Tr. 54:3–55:9. CSS did not develop or fund the publications that determined the tax bases for the counties. Auburn Dep. 166:24–165:8. No evidence was put forth to suggest the pricing methodology required CSS to invest significant resources. The Fourth Circuit has also determined that when a company only takes “minimal steps to ensure the confidentiality of its pricing information” that this information does not constitute a trade secret. IVS Hydro, 93 F. App’x at 528. The extent of the measures taken by CSS to guard the secrecy of its customer lists and prices do not rise to a level sufficient to consider this information a trade secret. Until 37 September 12, 2014, Christopher Herrington was not required to sign any type of confidentiality agreement. The evidence shows that other CSS employees were also not required to do so until around that time, despite CSS’s many years in business. It is not clear who had access to the pricing information internally. There was no evidence put forth that CSS’s client list was subject to any internal constraints or confidentiality restrictions. The Fourth Circuit has also found that when prices are “common knowledge in the marketplace” that trade secret protection is not available. IVS Hydro, 93 F. App’x at 528. Again, the prices charged by CSS are also publically available information subject to document requests pursuant to the West Virginia Freedom of Information Act. See W. Va. Code § 29B–1–2-(5). Furthermore, CSS has also failed to meet its burden of persuasion regarding misappropriation. First, CT has an entirely different pricing model from CSS. CSS sells its software as well as support services, while CT does not charge anything for the software, and instead charges only for support. Pl.’s Mot. Prelim. Inj., Ex. B at 108:19–109:2, 342:13–343:19 [ECF No. 65-3] (“C. Herrington Dep.”). Furthermore, the evidence put forth at the hearing suggests that Christopher Herrington became aware of or attempted to become aware of contract pricing by directly soliciting customers himself and by offering to provide the counties a better price. See, e.g., Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 41:16–43:1; Defs. Exs. 14, 15, 16 from Prelim. Inj. Mot. Hr’g [ECF Nos. 96-13, 96-14, 96-15]. Thus, I FIND that CSS is not likely to succeed on the merits of its WVUSTA 38 claim with respect to its pricing methodology. b. Geographic Expansion to Ohio CSS’s expansion to Ohio does not constitute a trade secret. I am skeptical that it is possible to obtain trade secret protection over the idea that a software solution provider to county governments might one day choose to expand to a neighboring state, after the company had already established business in that state. The WVUSTA requires that a trade secret be “the subject of efforts that are reasonable under the circumstances to maintain its secrecy.” W. Va. Code Ann. § 47– 22–d–2. CSS did not take many measures to keep this confidential. Many within the company were aware of this desire. Hr’g Tr. Feb. 23, 2017, 163:5–9. Additionally, this information could be acquired and duplicated very easily by others once CSS actually established a business presence in Ohio. Hr’g Tr., Feb. 23, 2017, C. Herrington Cross, 254:18–19 (“They [] had one account there.”); State ex rel. Johnson, 419 S.E.2d at 1. Most importantly, CSS has not been able to show that CT has misappropriated this information, even if it were considered a trade secret. All CSS has been able to show thus far is that Christopher Herrington has spoken to an individual with contacts in Ohio. Id. at 253:22–254:1. CT is not in Ohio nor does it appear to have any plans to expand into Ohio. Thus, I FIND that CSS is not likely to succeed on the merits of its WVUSTA claim with respect to expansion to Ohio. c. CSS’s Customers’ Preferences CSS alleges that knowledge it gained through its interactions with county 39 governments constitutes protectable trade secrets. CSS alleges that Christopher Herrington learned about these county practices in the years he worked at CSS. Auburn Dep. 156:18–159:20. I FIND that this information does not constitute a trade secret. It does not appear that CSS took any measures to keep this information confidential. Again, CSS presented no evidence that it ever limited the number of employees who could access this information or that CSS informed employees that the information was confidential. Regardless of whether or not this information may be considered a trade secret, evidence of disclosure or misappropriation by CT or Christopher Herrington was not presented. The type of information alleged largely has to do with whom to contact at the office. WVUSTA only prevents someone from acquiring a trade secret using “improper means.” W. Va. Code § 47–22–1. It does not preclude someone from going out and developing that information on his own. From the evidence put forth, it is clear that Christopher Herrington was able to obtain this information on his own by contacting county clerks on a list provided by the West Virginia Association of Counties, developing relationships, and asking them to provide feedback about their preferences. See C. Herrington Dep. at 50:3–11. Therefore, I FIND that CSS is not likely to succeed on the merits of the WVUSTA claim as related to CSS’s customer preferences. C. Likelihood of Success on the Merits for Breach of Contract The likelihood of CSS prevailing on the merits of its breach of contract claim 40 depends on whether the court will enforce the provisions of the Confidentiality Agreement.16 To analyze whether those provisions are enforceable, the court first must determine which state’s law to apply. The court must employ West Virginia’s choice-of-law principles because this court sits in diversity in West Virginia. See Klaxon Co. v. Stentor Elect. Mfg. Co., 313 U.S. 487, 496 (1941). 1. Choice of Law The Confidentiality Agreement contains a choice of law provision, which identifies Texas law as controlling. Confidentiality Agreement ¶ 7 [ECF No. 65-2]. Nevertheless, no signatory to the Confidentiality Agreement is from Texas, the agreement was entered into in West Virginia, and the alleged breach occurred in West Virginia. Under West Virginia’s conflict of laws jurisprudence, “[a] choice of law provision in a contract will not be given effect when the contract bears no substantial relationship with the jurisdiction whose laws the parties have chosen to govern the agreement.” See Gen. Elec. Co., 275 S.E.2d at 290. The parties agree that West 16 The relevant portion of the Confidentiality Agreements states the following: “Confidential and Proprietary Information” means any and all data and information which is (1) disclosed to or known by Employee as a consequence of, during, or through Employee’s association with the Company; (2) not generally known outside the Company; and (3) which relates to any aspect of the Company or its business, research or development. Confidential and Proprietary Information includes, but is not limited to: research, technology, strategic plans, business plans, information relating to the Company’s operating plans and methodologies, operational manuals, pricing and marketing strategies, financial information, and other trade secrets, as well as any or all information which is disclosed to Employee or in any way acquired by Employee relating to the Company or its employees, clients, customers, prospective customers or clients, suppliers, or vendors. Confidential and Proprietary Information may be written, verbal or recorded by electronic, magnetic or other methods, whether or not expressly identified as “Confidential” by the Company. Confidentiality Agreement 1 [ECF No. 65-2]. 41 Virginia law applies.17 Pl.’s Mem. Supp. Mot. 5, 6 (“The non-copyright claims of CSS are generally governed by West Virginia law. . . . The Confidentiality Agreement was formed and performed in West Virginia and, therefore, its valid[ity] would generally be governed by West Virginia law.”). Thus, I FIND that West Virginia law should control the court’s interpretation of the Confidentiality Agreement. 2. Validity of the Confidentiality Agreement CSS alleges that Christopher Herrington breached the Confidentiality Agreement. First Am. Compl. ¶ 43 [ECF No. 59]. Specifically, CSS argues that Christopher Herrington must be in breach of the agreement because he would not be able to service his new clients without using detailed knowledge of the plaintiff’s software code acquired during his employment. Pl.’s Mem. Supp. Mot. 8. The plaintiff states that Christopher Herrington breached the agreement by “(1) using the Confidential Information to create and market and sell systems to CSS’s customers and (2) by using the Confidential Information to sell maintenance services for CSS’s software.” Id. at 9. The plaintiff argues that the Confidentiality Agreement is reasonable because the agreement applies only to “information not generally known outside the company which relates to CSS’s business, research, or development,” and “there is no preclusion on [Christopher Herrington]’s use of general programming skills or services.” Id. at 8. The defendants assert that the agreement is too broad, rendering it unenforceable under West Virginia’s “Rule of Reason” test, because “it encompasses essentially all the information Christopher Herrington was provided by 17 CSS argues that it would also be successful if Texas law were to be applied. 42 CSS or learned during the course of his employment with CSS.” Defs.’ Mem. 4 [ECF No. 68]. The West Virginia Supreme Court of Appeals has not squarely addressed the enforceability of a non-disclosure agreement. Nevertheless, I have previously held that non-disclosure agreements should be examined under the same legal structure as non-compete agreements. See McGough, 496 F. Supp. 2d at 752. I held that “[c]ovenants not to disclose and utilize confidential business information are related to general covenants not to compete because of the similar employer interest in maintaining competitive advantage.” Id. at 756. In deciding whether restraints on disclosure are reasonable two factors are important: (1) whether the employer is attempting to protect confidential information related to the business, such as trade secrets, methods of operation, names of customers, and personnel data even though the information does not rise to the status as a trade secret, and (2) whether the restraint is reasonably related to the protection of the information. Id. (citing Restatement (First) of Contracts §§ 515(a), 516 (Am. Law Inst. 2017)). The framework to apply to non-disclosure agreements is as follows: 1. 2. 3. 4. Is the agreement a binding contract at law? Is the agreement unreasonable on its face? Does the employer have an interest requiring protection? Is the employee able to rebut the presumption of enforceability? See Reddy v. Cmty. Health Found. of Man, 298 S.E.2d 906, 915–17 (W. Va. 1982) Upon close examination of the Confidentiality Agreement at issue here, I FIND that it is a properly formed contract. 18 I also FIND that the contract is not 18 Defendants do not challenge the confidentiality agreement on the grounds of being an improperly formed contract. Hr’g Tr., Jan. 25, 2017, 22:7–23. 43 unreasonable on its face. In fact, what the Confidentiality Agreement covers is narrow. To be considered “confidential information” under the agreement, three conditions must be satisfied: (1) Christopher Herrington used information acquired through his employment that (2) was not generally known outside of the company and (3) that the information related to the plaintiff’s business, research, or development. Whether CSS is likely to succeed on the merits of the breach of contract claim depends on whether it can show that Christopher Herrington divulged confidential information as defined by this agreement. As discussed in previous sections of this opinion, CSS has not done so. Accordingly, I FIND CSS is not likely to succeed on the merits of this claim. D. CSS Will Not Suffer Irreparable Harm Plaintiffs seeking preliminary relief must demonstrate that irreparable injury is likely in the absence of an injunction. Winter, 555 U.S. at 22; see also 11A Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2948.1 (2d ed. 1995). A “possibility of irreparable harm” is not enough to earn a preliminary injunction. Real Truth About Obama, 575 F.3d at 346. Plaintiffs argue that without this injunction they will suffer in many ways; however, CSS has not shown it will be irreparably harmed before trial absent preliminary injunctive relief. Moreover, on December 16, 2016, CSS moved this court for an extension to a reply to its Motion for Preliminary Injunction [ECF No. 68]. This further persuades me that CSS is not 44 likely to face irreparable harm. CSS is concerned that CT will use its IP to become a permanent competitor. Because CSS is not likely to succeed on the merits of its copyright claim or trade secret claims, this is not a convincing argument. Courts are not in the business of removing a plaintiff’s competitors from the marketplace because the plaintiff is losing business. See Robie v. Price, No. 2:17-CV-03089, 2017 WL 3097529, at *6 (S.D. W. Va. July 20, 2017) (“[T]he sole fact that a company is losing money does not irreparable harm make.”). In addition to CT becoming a permanent competitor, CSS worries simultaneously CT will no longer be a viable company. That a party may go out of business as a consequence of litigating a suit is a danger present in any case. There is also no guarantee that CSS will be able to recover any monetary damages, even should it receive a final judgment in its favor. Hr’g Tr., Feb. 24, 2017, 67:1–4. This is a risk that CSS knowingly assumed when it initiated this action. I am not convinced that monetary damages would be an insufficient remedy in this case. CSS alleged damages in the form of lost profits, future lost profits, and future potential earnings from “bundled sales”19—upselling their current customers on new software offerings. These harms all relate to monetary losses. However, CSS has not provided any tangible monetary damages or estimates regarding these potential future earnings or losses. CSS says it is impossible to prove 19 In the twenty-five year history of the company prior to Christopher Herrington’s 2015 departure, CSS had regularly made additional “bundle sales.” Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 127:9–15. Since CT emerged into the market, CSS has made virtually no bundled sales to its customers. Id. at 127:20–24. 45 these exact losses with any substantial certainty and that some of its alleged damages are difficult to calculate. Hr’g Tr., Feb. 23, 2017, D. Herrington Direct, 127:25–128:21; Hr’g Tr., Jan. 25, 2017, 63:4–7. CSS has sufficient financial information regarding its own finances and contracts to calculate damages readily for the value of customers lost to CT. CSS should also be able to calculate its actual damages for its copyright claim. Actual damages include those suffered the copyright owner “as a result of the infringement, and any profits of the infringer that are attributable to the infringement.” 17 U.S.C. § 504(b) (emphasis added). There is a dispute as to whether the alleged infringement occurred before or after the effective date of the copyrights. In either case,“[a]ctual damages are proper for acts of infringement occurring both before or after the effective registration.” 17 U.S.C. § 504(b). CSS also alleges that its reputation and goodwill have been damaged as a result of this litigation. Hr’g Tr., Jan. 25, 2017, 63:17–22; Mem. Law Supp. Pl.’s Mot. Prelim. Inj. at 17. However, CSS did not present any evidence of this. An expert may have been able to opine on whether or not this is connected, but it has not been made clear this is the case. Thus, I FIND that CSS has not made a showing that it will likely suffer irreparable harm absent a preliminary injunction. E. Balancing of Hardships Does Not Tip in CSS’s Favor A plaintiff seeking a preliminary injunction must establish that the balance of equities tips in his favor. Winter, 555 U.S. at 20. CSS argues that it would be harmed 46 by continued infringement of its intellectual property. Pls. Mot. Prelim. Inj. 18. CSS also asserts that nothing in the Confidentiality Agreement would have prohibited Christopher Herrington from working on programs in an unrelated interest. Id. CT expresses concerns that it will be unable to stay in business. Defs. Mem. Opp. Pls. Mot. Prelim. Inj. 19. CSS has not provided sufficient evidence to show it is likely to succeed on this factor. Therefore, I FIND that CSS has not made a showing that the balance of hardships weighs in its favor. F. An Injunction is Not the Public Interest CSS asserts that granting its motion for a preliminary injunction is in the public interest because the Fourth Circuit has recognized that “the public has an interest in enforcing restrictive covenants that protect business interests, and that the public has an interest in preventing . . . infringement of copyrights.” Bowe Bell, 145 F. App’x at 404. “Since Congress has elected to grant certain exclusive rights to the owner of a copyright in a protected work, it is virtually axiomatic that the public interest can only be served by upholding copyright protections and, correspondingly, preventing the misappropriation of the skills, creative energies, and resources which are invested in the protected work.” Klitzner Indus., Inc. v. H. K. James & Co., 535 F. Supp. 1249, 1259–60 (E.D. Pa. 1982) (emphasis added). However, “[c]opyright law seeks to ‘balance the competing concerns of providing incentive to authors to create and of fostering competition in such creativity.’” Daniel A. Crowe, The Scope of Copyright for Non-Literal Design 47 Elements of Computer Software: Computer Associates International, Inc. v. Altai, Inc., 37 St. Louis U. L.J. 207, 233 (1992). If copyright protection “is extended too far, it will impede technical innovation by not allowing other developers in the computer industry to build upon and expand the work of others.” Id. There is not yet a brightline test to determine the scope of software copyright protection, and thus, this task has been left to the courts. Id. Given that CSS is not likely to succeed on the merits of its copyright claim or trade secret claims, this is not a compelling argument for the plaintiff. The Fourth Circuit also encourages a court to consider the public interest in making a decision on whether to issue a preliminary injunction. Adkins v. Scott, 597 F.2d 872, 877 (4th Cir. 1979). The Supreme Court instructs courts to “pay particular regard for the public consequences in employing the extraordinary remedy of injunction.” Winter, 555 U.S. at 24. CSS asks for CT be estopped from providing support services to its current customers: twenty clerks’ offices in the State of West Virginia. See Hr’g Tr., Feb. 24, 2017, C. Herrington Direct, 43:19–44:5. Bert Auburn states that CSS would be able to provide support services for CT’s clients before switching them back to CSS’s software. Hr’g Tr., Feb. 24, 2017, Auburn Direct 68:17– 69:3. However, an injunction would no doubt negatively disrupt the operation of the county clerks’ offices during such a transition. Thus, I FIND that CSS has not shown that the injunction is in the public interest. VI. Conclusion 48 For the reasons outlined above, the Court ORDERS that Plaintiff’s Motion for Preliminary Injunction [ECF No. 65] is DENIED. The Court DIRECTS the Clerk to send a copy of this Order to counsel of record and any unrepresented parties. ENTER: August 3, 2017 49

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.