Re: You have rights.. Software Patents are Illegal

> Computer programs are subject to copyright, but that doesn't mean that
> the function that they perform cannot be subject to patent. ... There
> is no real difference between translating a computer program into an
> executable image in memory, and transforming a design into a working
> machine.

Can't fault that.  However - though I didn't entirely follow Christian's
argument - as I understand it, the latest proposals of the EU oligarchs
would make it an offence to publish a program (in source-code form)
which implements an idea covered by a patent.  If I draw up a blue-print
for a machine which (if anyone made it) would be covered by your patent,
and my blue-print is independently created, I would not be infringing
either your patent or your copyright in *your* blue-print.  Software
would appear to be getting the worst of both worlds ...

Part of Christian's argument is that the author of software should - by
the laws covering literary works - be guaranteed full liberty to
distribute that copyright work, just as I am fully at liberty to write
an academic paper (or journalistic article) describing any patented
idea, as long as I do so in my own words, without copying from others.
I don't know how far that reasoning can be dragged, but it does point to
an interesting issue - if the compiled program is a machine that can
infringe a patent, then it can't be a literary work, in which case the
software industry loses the ability to restrict copying of compiled code
(i.e. you can give your friends copies of your operating system even if
it isn't open source - albeit you can't distribute the source without
the author's consent).

If I write a program in pseudo-code that outlines an algorithm, my
writing is expressive, so protected by copyright; to claim that it
infringes a patent, you must show how the pseudo-code performs the
computation; which it doesn't, because we haven't compiled it.  However,
that distinction doesn't work, because there are interpreted languages
which are designed to be expressive - the source is what the computer
runs.  However, since it's *performing* the patented acts that is
covered by a patent, not *describing the means to do so*, there is a
strong case that source code should never be construed to infringe a
patent, even if anyone who runs it does infringe.

So I *think* Christian's logic is that there's a collision between the
copyright treaties' insistence that authors have whole authority over
their work and, if patents are granted on software, the patent treaties'
insistence that an implementation of an idea, even if independently
arrived at, falls under the control of whoever patented it.

I'm not sure how far that can be taken, but I'm interested,

	Eddy.