FEDERAL COURTS LAW REVIEW -- 1998 Fed. Cts. L. Rev. 1
Discovery in Computer Software PatentLitigation
By Andy Johnson-Laird
Abstract
[a.1] One of the more
frustrating aspects of computer software patent litigation is that
discovery productions are usually demonstrably incomplete, and the
information that is produced is produced in a maximally
inconvenient form. These frustrations flow from a series of
assumptions -- false assumptions, as this article explains. The
false assumptions are as follows:
a)
Source code, the human-readable form of software, appears to
the untrained eye as text. Therefore, some assume that a printed
listing of source code is an appropriate response to a document
production request.
b) Source
code exists as a series of self-contained text files (or so it
is assumed) and therefore, some assume that if all of these
text files are printed out, the entirety of the source code will be
produced.
c) Object code, the
computer-readable form of software, if printed out, is assumed to
constitute a legitimate production to the requesting
party.
All three assumptions are
false. The false assumptions lead to incomplete discovery
productions.
[a.2] The analysis of source code on
paper is far more costly in time and money than is the analysis of
that same material on computer media.(1)
Simple searching of thousands of lines of source code on paper may
take hours, in contrast to a few seconds were that source code on
computer media. Printed listings of source code are usually
incomplete, although it may take several days or weeks of analysis
to detect that.
[a.3] Source code printed on paper lacks
important forensic information contained in that same source code on
computer media. Printed object code is essentially useless when
compared with that same object code on computer media.
[a.4] The
absolute truth about a computer program's behavior resides in one
place and one place alone: the computer program as it executes on a
computer. The next nearest representation of the absolute truth is a
static copy of that program's object code on computer media,
followed, in order of increasing distance from the absolute truth,
by the computer source code on computer media and the computer
source code on paper. Object code on paper is almost worthless
except for the analysis of very small pieces of the
program.
[a.5] Producing parties who avoid these
false assumptions and produce relevant computer software on
diskettes or magnetic tape usually presume that it will be possible
to divine how this computer media can be accessed to reload the
information it contains. That presumption is usually false. To
access the information, one needs to know very specific information
about the hardware and software used to write the relevant software
onto the computer media.