Re: GPL 2(b) HUH?

[Rjack]

Ciaran O'Riordan wrote:

> Whether X is a derivative of Y is determined at time of writing, not 
> at time of linking or compiling.

Not true.

"1 Nimmer on Copyright § 3.02 ('[T]he originality called for in a
collective work consists of the collection and assembling of
pre-existing works, while derivative work originality lies in the
manner in which a preexisting work is transformed . . . .')."

From Nimmer (supra):

"... derivative work originality lies in the manner in which a
preexisting work is transformed"

So, the *manner* in which the transformation takes place determines
whether a derivative work is created.

A one to one mapping from one set of symbols (source code) symbols into
a second set of symbols (machine code) with an existing, predetermined
algorithm (gcc) contributes no creative originality to the work since
anyone of millions of people who pushe the button invoking gcc get an
identical result. (I certainly didn't write gcc so I have no creative
contribution there.)

A one to one mapping from one set of symbols into a second set of
symbols is just paraphrasing the general definition of an algorithmic
mathematical function.

If I have permission to copy your source code and I run a gcc
"translation" on it, I have created a new copy and not a new derivative
work.

Linux consists of a collection of hundreds of independently copyrighted
works (source code modules) -- it is truly a "collective work".

If I take those independent works add a couple of my own independent
works and re-arrange the source tree I have a created a new collective work.

If I write a new Makefile and compile and link the new source tree with
gcc I have *not* created a new derivative work -- simply an unoriginal
algorithmic translation of an existing collective work.

Sincerely,
Rjack :)