Post Syndicated from Bradley M. Kuhn original http://ebb.org/bkuhn/blog/2007/05/08/infringement.html

Nearly all software developers know that software is covered by
copyright. Many know that copyright covers the expression of an idea
fixed in a medium (such as a series of bytes), and that the copyright
rules govern the copying, modifying and distributing of the work.
However, only a very few have considered the questions that arise when
trying to determine if one work infringes the copyright of
another.

Indeed, in the world of software freedom, copyright is seen as a system
we have little choice but to tolerate. Many Free Software developers
dislike the copyright system we have, so it is little surprise that
developers want to spend minimal time thinking about it.
Nevertheless, the copyright system is the foremost legal framework
that governs software¹, and we have to
live within it for the moment.

My fellow developers have asked me for years what constitute copyright
infringement. In turn, for years, I have asked the lawyers I worked
with to give me guidelines to pass on to the Free Software development
community. I’ve discovered that it’s difficult to adequately describe
the nature of copyright infringement to software developers. While it
is easy to give pathological examples of obvious infringement (such as
taking someone’s work, removing their copyright notices and
distributing it as your own), it quickly becomes difficult to give
definitive answers in many real world examples whether some particular
activity constitutes infringement.

In fact, in nearly every GPL enforcement cases that I’ve worked on in
my career, the fact that infringement had occurred was never in
dispute. The typical GPL violator started with a work under GPL, made
some modifications to a small portion of the codebase, and then
distributed the whole work in binary form only. It is virtually
impossible to act in that way and still not infringe the original
copyright.

Usually, the cases of “hazy” copyright infringement come up
the other way around: when a Free Software program is accused of
infringing the copyright of some proprietary work. The most famous
accusation of this nature came from Darl McBride and his colleagues at
SCO, who claimed that something called “Linux” infringed
his company’s rights. We now know that there was no copyright
infringement (BTW, whether McBride meant to accuse the GNU/Linux
operating system or the kernel named Linux, we’ll never actually
know). However, the SCO situation educated the Free Software
community that we must strive to answer quickly and definitively when
such accusations arise. The burden of proof is usually on the
accuser, but being able to make a preemptive response to even the hint
of an allegation is always advantageous when fighting FUD in the court
of public opinion.

Finally, issues of “would-be” infringement detection come
up for companies during due diligence work. Ideally, there should be
an easy way for companies to confirm which parts of their systems are
derivatives of Free Software systems, which would make compliance with
licenses easy. A few proprietary software companies provide this
service; however there should be readily available Free Software tools
(just as there should be for all tasks one might want to perform with a
computer).

It is not so easy to create such tools. Copyright infringement is not
trivially defined; in fact, most non-trivial situations require a
significant amount of both technical and legal judgement. Software
tools cannot make a legal conclusion regarding copyright infringement.
Rather, successful tools will guide an expert’s analysis of a
situation. Such systems will immediately identify the rarely-found
obvious indications of infringement, bring to the forefront facts that
need an exercise of judgement, and leave everything else in the
background.

In this multi-part series of blog entries, I will discuss the state of
the art in these Free Software systems for infringement analysis and
what plans our community should make for the creation Free systems
that address this problem.

---

¹ Copyright is the legal
system that non-lawyers usually identify most readily as governing
software, but the patent system (unfortunately) also governs software
in many countries, and many non-Free Software licenses (and a few of
the stranger Free Software ones) also operate under contract law as
well as copyright law. Trade secrets are often involved with software
as well. Nevertheless, in the Software Freedom world, copyright is
the legal system of primary attention on a daily basis.