By Jason Stiehl

Last month, Judge Walls of the U.S. District Court of New Jersey became yet another pioneer in the evolving world of trade secret protection and the Internet. In a well-reasoned and thorough analysis of case precedent, Judge Walls utilized two historic landmark public disclosure cases, DVD Copy Control Ass’n, Inc. v. Bunner, 116 Cal. App. 4th 241 (Cal. App. 6th Dist. 2004), and Data General Crop. v. Digital Computer Controls, Inc., 357 A.2d 105 (Del. Ch. 1975), as his guidepost in determining whether certain code language found within Syncsort’s Reference manual and scripts remained trade secrets, despite posting of both parts, and, in some instances, all of the language on various websites.

The Guiding Hand of History

Data General and DVD Copy present a clear contrast of partial disclosure versus unlimited disclosure. In Data General, a pre-Internet disclosure case, a minicomputer manufacture made publicly available, through manuals, general technical information governing its products. The Court held that, unlike a logic diagram, the manuals did "not contain sufficient logic design [ ] to permit their being successfully used of the purpose of either duplicating such machine or in assembling a computer substantially identical to" the Data General’s minicomputer. Data General, 357 A.2d at 110-11. In contrast, in DVD Copy, a foreign computer programmer, through his license agreement, began widespread distribution of the DIVX code associated with copyright protection on DVDs. Rather than suing this individual, the association tasked with protecting the copyright status on DVD’s went after a host of United States’ individuals who had posted portions of the code on their websites and blogs. Ultimately, by the time the matter came to preliminary injunction, the information had been distributed to over a million people, thus, according to the court, eviscerating the trade secret status of the code, holding that information "in the public domain cannot be removed… under the guise of trade secret protection." DVD Copy, 116 Cal. App. 4th at 255.

Bytes and Pieces

In Syncsort, the Plaintiff developed a language which allowed users to translate data from one form to another. Syncsort’s competitor, Innovative Routines, Int’l ("IRI"), also maintained proprietary software which allowed for a similar translation. However, because the languages were unique, it was difficult and time-consuming for a Syncsort customer to simply switch to IRI’s program. To solve this problem, IRI improperly, through a Syncsort distributor, came into possession of a Syncsort Reference Guide which contained over 400 pages of description and definition related to Syncsort’s language. IRI then took this guide and developed a translator for Syncsort’s translation device– called ssu2scl– which could translate Syncsort scripts to IRI’s program language. IRI also requested, from Syncsort customers, examples of Syncsort scripts to run against the ssu2scl to determine the effectiveness of the translation device. The activities of both the distributor and the customers were controlled and governed by confidentiality and licensing agreements prohibiting such disclosures.

Once sued for its activities, IRI went about a hunt throughout the Internet to locate Syncsort’s Reference Guide and scripts to demonstrate that, although they may have procured the information from an improper source, the information was publicly available, and therefore should not be afforded trade secret status. Mostly unsuccessful, IRI was initially able to locate only four partial sources and, ultimately, three full sources, where Syncsort’s information was available. As to the partial sources, Judge Walls applied the logic of Data General and held that the information, in its fragmented and limited form, was not sufficient to recreate the Syncsort language. As to the full sources, Judge Walls looked to DVD Copy, and, in contrast, found that the full posts– found on (1) a university password-protected site; (2) a Korean website taken down within days of notice; and (3) a Japanese website taken down within days of notice– were "sufficiently obscure of transient or otherwise limited" so that it was not "generally known to the relevant people."

A "Manual" Going Forward?

Although it will be tempting for litigants to cite this case for a black-letter type pronunciations, this author would caution against such efforts. First, the facts of this case tilted well against the Defendant. For example, IRI had: (1) admittedly improperly sought out and received the information; (2) only conducted internet searches after being sued to determine whether the information was publicly available; and (3) known that the sources of the information were bound by restrictions governing the sharing of that information. Second, this case has a prolonged and protracted history, including a previous trial on the merits and full summary judgment briefing, allowing for a complete record to develop. Third, the breadth and depth of the release of information remains a case-by-case type determination without any precise formulation. For example, imagine if the Korean manual had been up for five years, or if it had been translated from Korean and posted on an US website. With that said, this case presents what will likely become the paradigm for Internet "release" cases in the future.