Re: The death of copyright in software

[Stefaan A Eeckels]

On Thu, 31 May 2007 20:19:17 -0500
rjack <rjack@com> wrote:

> Run your C++ compiler code through the
> “abstraction-filtration-comparison” test in the hands of an expert
> witness in court and your source modules look like Swiss cheese with
> VERY large holes. If the programmer’s comments have been stripped
> (very likely) and trivial obfuscation steps have been applied, your
> copyright protection is virtually non-existent.

If you strip out all the distinguishing characteristics, it's pretty
obvious that on the remainder you cannot get copyright protection. If
you strip the distinguishing stuff out of the typical novel, the result
is no longer copyrightable either (e.g. Woody Allen's appreciation of
War and Peace: "It's about Russians", or a reduction of Romeo and
Juliet to "Boy meets girl, they fall in love, families don't want
them to marry", etc). If you reduce the number of notes you look at to
three or four, every piece of music is a copy of every other piece of
music. 

It's obvious that you cannot claim copyright protection on fragments of
code that anyone would write substantially the same because these
fragments are determined in their expression by the hardware or
other external influences.

But only an idiot without knowledge about programming can argue that
because a program performs the same well-defined function as another
program (i.e. compiling 'C' code or performing an FTP transfer) its
internal structures and algorithms have to be so similar as to be
indistinguishable "after applying trivial obfuscation".

Using your overly broad approach you could equally claim that love
stories cannot be copyrighted, or that pictures of the Eiffel tower
cannot be copyrighted because when you strip off all the distinguishing
characteristics (framing, lighting, etc) they are pictures of the
exact same object. But the reality is that every photograph is
copyrighted, no matter how similar or trivial the subject. You're free
to make your own picture of the scene, but you're not free to copy
someone else's picture. The same applies to source code.

What the Lexmark case is about is that similarity between those parts of
the code that are so determined by the function to be performed cannot
be used in court as proof of copyright infringement. Quite obviously it
does not mean that you can lift sections out of the gcc source code and
use them in your own compiler because "there is only one way in which a
compiler can be written" and hence you can do with the gcc code as you
please.

-- 
Stefaan A Eeckels
-- 
   The one thing IT really needs to outsource is the freakin' clueless
 managers that don't understand that there are more possibilities than
chaos on the one hand and the reduction of alternatives to zero on the
other.                    -- Richard Hamilton in comp.sys.sun.hardware