Discovery of Source Code in Patent Litigation
Discovery of computer source code—either through production, inspection, or deposition—is one of the more contentious aspects of patent litigation. Indeed, “few tasks excite a defendant less . . . . Engineers and management howl at the notion of providing strangers, and especially a fierce competitor, access to the crown jewels. Counsel struggle to understand even exactly what code exists and exactly how it can be made available for reasonable inspection. All sorts of questions are immediately posed. . . . Put simply, source code production is disruptive, expensive, and fraught with monumental opportunities to screw up.” Apple Inc. v. Samsung Elecs. Co., No. 11-1846, 2012 U.S. Dist. LEXIS 62971, *10-11 (N.D. Cal. May 4, 2012) (ECF No. 898).
On one hand, the patentee wants the code to particularly show or confirm that a product or service infringes (or operates in a certain manner) rather than pointing to technical documents (e.g., manuals, diagrams, etc.) which, though helpful, may not have the same specificity that code provides. The converse could also be true where the code’s meaning is debatable, while the technical documents explain in detail the accused system. On the other hand, the accused infringer is deeply concerned that the source code—which is the foundation of its products and thus represents a significant monetary investment—may seep to a competitor and accordingly alter the competitive landscape.
Recent cases generally have ordered source code be made available because it is relevant given the broad discovery permitted in litigation and it could be secured through a protective order or confidentiality agreement. Yet, because source code is so sensitive, courts have sometimes denied production where the code is duplicative of other information produced or where the protective order is insufficient in light of the competition between parties.
Production of Code
Source code is discoverable if the defendant fails to meet its burden by providing specific objections. For example, in Marketlinx Inc. v. Industry Access Inc., et. al., No. 12-03496 (C.D. Cal. Jan. 2, 2013) (Olguin, MJ.) (ECF No. 84), the Court compelled the defendant, Instinet, to produce source code for each version of the accused product dating back to June 2005. The defendant argued that the plaintiff never demonstrated the relevancy of code before the patent issue date (i.e., March 2008) and generally objected to discovery as “overly broad” and “unduly burdensome.” The Court viewed the latter argument as conclusory, and dismissed the former because it is the defendant’s (not the plaintiff’s) burden to defeat a discovery request and regardless “courts generally allow discovery to extend to events before the period of actual liability so as to provide context.”
To protect the confidential or possible trade secret nature of source code, courts will look to any protective order or confidential agreement in place and whether technical documents and other materials can substitute for the code. In Impulse Tech. Ltd. v. Nintendo of America, Inc., et. al., No. 11-02519 (N.D. Ohio Sept. 19, 2012) (ECF No. 120), the Court granted plaintiff’s motion to compel because, although the security of the source code is a “weighty concern,” the confidentiality agreement entered into previously between the parties was sufficient. But, not all courts are likewise convinced. The Court in MagicJack Vocaltec Ltd. v. NetTalk.com, Inc., No. 12-80360 (S.D. Fla. Oct. 18, 2012) (ECF No. 51) denied plaintiff’s motion to compel source code without prejudice because the Court was not “persuade[d] . . . that the source code is necessary to fully analyze Defendant’s accused devices” based on dueling declarations filed, and did not think the agreed upon protective order provided enough protection or an “impregnable wall” and thus believed disclosure could be fatal to defendant’s business. See also Generac Power Sys. Inc. v. Kohler Co., No. 11-1120 (E.D. Wisc. Jun. 6, 2012) (ECF No. 32) (“Source code for programs is often one of the most valuable assets a company possesses.”); MVS Studio Inc. et al v. Bingo Bean LLC et al., No. 10-07675 (C.D. Cal. Jun. 24, 2011) (ECF No. 73) (Segal, MJ.) (permitting the production of source code with the agreed upon caveat that any party who violates the protective order will dismiss its claims with prejudice).
Technical documents and other materials may be sufficient to prevent discovery of source code. The Court in Generac Power denied a motion to compel source code because it was concerned about the code being used by the Plaintiff to improve its products and believed that other substitutes or reasonable accommodations were provided by the defendant, including software demonstration and manuals, which in the Court’s opinion would provide the same information. But see Edward D. Ioli Trust, et. al. v. Avigilon Corp., et. al., No. 10-00605, 2012 U.S. Dist. LEXIS 164425, *5-10 (E.D.Tex. Nov. 16, 2012) (ECF No. 279) (rejecting defendant Vigilant’s local patent rule interpretation that it could produce technical documents rather than source code and noting that it is not the plaintiffs’ burden to identify any deficiencies in defendant’s document production).
Even where the source code is treated as a trade secret, it may be discoverable with a particularized showing. In Fleming v. Escort, Inc., No. 09-105, 2010 U.S. Dist. LEXIS 101938, *4-6 (D. Idaho Sept. 24, 2010) (ECF No. 78) the defendant produced redacted source code claiming the unredacted portions relate to non-infringing products. In opposition, the plaintiff submitted a declaration from a professor claiming that the unredacted portions were necessary to understand the code. This persuaded the Court despite the code being “obviously highly sensitive trade secret material . . . .” See also Nazomi Commcns., Inc. v. Samsung Telcoms., Inc., No. 10-05545, 2012 U.S. Dist. LEXIS 76468, *10-12 (N.D. Cal. Jun. 1, 2012) (ECF No. 192) (“Given the sensitivity of source code, the Court is not inclined to order broad disclosure absent a more specific showing.”) (citation omitted).
Courts have also discussed how the source code should be produced, including weighing the costs of production. In ProconGPS, Inc. v. Star Sensor, LLC, No. 11-03975, 2012 U.S. Dist. LEXIS 154868, *3-4 (N.D. Cal. Oct. 29, 2012) (ECF No. 92) the Court ordered the defendant to produce source code in a format that is “searchable, hierarchical, or contained developer notes” but only “[t]o the extent that the source code was originally maintained in” that format. Defendant originally produced 28,000 pages on a disc and thus made it harder for the plaintiff to review the code. See also MedioStream, Inc. v. Microsoft Corp., No. 8-369, 2010 U.S. Dist. LEXIS 110420, *20-22 (E.D. Tex. Oct. 18, 2010) (Everingham IV, MJ.) (ECF No. 468) (requiring code be produced in intelligible format, but Microsoft is not required to invest a “great expense” (i.e., millions of dollars) to duplicate its code repository).
Inspections and Depositions
Inspections of and depositions on source code are no less discoverable. In MobileMedia Ideas LLC v. HTC Corp., et. al., No. 10-00112 (E.D. Tex. Oct. 19, 2012) (ECF No. 211), the Court ordered the defendants, HTC Corporation and HTC America, Inc., to print 7,700 pages of its source code. The parties had agreed in a protective order that source code would be made available for inspection with the defendants agreeing to print any pages requested. MobileMedia as part of its inspection identified 1,000 out of 4,000,000 source files (or approximately 7,700 pages) to print. HTC refused because its “crown jewel” could inadvertently be disclosed and because all those pages would not be used at trial. The Court dismissed the arguments because discovery is broader than whether the evidence will ultimately be admitted at trial and MobileMedia represented that its print requests are narrowly tailored and will protect the code per the protective order.
Depositions on source code are “standard fare in patent cases.” Vasudevan Software, Inc. v. MicroStrategy, Inc., No. 11-06637 (N.D. Cal. Aug. 22, 2012) (Grewal, MJ.) (ECF No. 86) (ordering deponent to testify on the source code produced); Nomadix, Inc. v. Hewlett-Packard Co., et. al., No. 9-08441 (C.D. Cal. Jun. 21, 2011) (Kenton, MJ.) (ECF No. 395) (discussing, in a tentative decision, security measures in using source code at deposition, such as taking the deposition where the source code is located or where a secure laptop has been transported by a bonded carrier), discussed in (C.D. Cal. Jun. 28, 2011) (Kenton, MJ.) (ECF No. 397).
A source code request is a major milestone in patent litigation that cannot be taken lightly by either party. For accused infringers, they should first investigate whether they have documents and other information which would basically make the code superfluous. Also, the alleged infringers should research whether under state law the code is considered a trade secret, as in some jurisdictions, this may shift the burden of persuasion for source code production to the other party.
Next, alleged infringers should negotiate a protective order outlining the code’s scope (e.g., date, accused products only, versions, change logs, portion of the code only), availability (e.g., location, people, and number and length of reviews), format (e.g., native, paper or electronic form), uses (e.g., depositions), bates number procedure, and disclosure (e.g., penalties, logging and claw back) to ensure the confidentiality of its highly valuable source code and to ensure that accused infringers are not overly burdened in terms of cost and time.
Meanwhile, the patentee has its own considerations in seeking source code, such as using narrowly tailored requests and using a technical expert to explain in detail why certain code is necessary despite documents previously produced. This latter part is important in those jurisdictions that shift the burden if the code is given trade secret status.
In short, parties on either side of the “v.” must be fully cognizant of the considerations and issues surrounding source code discovery.