Subject: Re: [FYI] Microsoft license spurns open source
From: Lynn Winebarger <owinebar@free-expression.org>
Date: Tue, 26 Jun 2001 03:00:46 -0500

On Monday 25 June 2001 23:42, you wrote:
> On interesting difference that this may be bringing up is the difference
> between "code as speech" and "code as machine" (for lack of a better term)

   I believe "code as device" follows the patent language more closely.

> companies do not like to patent code 
    You could have fooled me.

> because then they would have to make
> it public (as part of the patenting process). Patenting speaks to the
> "code as machine" aspect of code.
>
    Actually, they don't have to fully disclose the code, only enough details 
of the algorithm so that a practitioner in the field can construct the 
device.  In particular, they can leave out details that would be required to 
make two instances of the class of devices described by the patent 
interoperable.

> "Fair use", "copyright", "free like speech", etc. speak to the "code as
> speech" part of code: i.e. if I print out DeCSS on a T-shirt, can I be
> liable in the same way as if I ran an illegal CD press?
    If you mean in terms of copyright infringement, yes. 

> The important distinctions between the "speech" and "machine" code model
> crop up in these use discussions. The GPL allows all changes to be kept
> proprietary 
   ^^^^^^^^^
  I think you mean private. 
> Closed and proprietary. I can even run proprietary apps ontop on a
> modified GPL software without having to release my modified code (as in
> the case of Apache, i beleive). Note: this is not part of "fair use"
> provisions in the copyright law, this is part of the freedom the GPL
> gives me.

    Actually, it can be construed as a result of the first sale doctrine.

> So, according to this, in Seth's code example, you would be free to study
> and modify the code without Seth's permission so long as you did not
> redistribute, and include excerpts (even modified) in certain
> nonexecutable forms (under fair use) that you could redistribute.

    What makes you think non-executable is important?  Copyright explicitly 
extends no protection to purely functional aspects of software.  

> Ownership confers more rights than liscensing, especially under UTICA.
> This is why computer companies do not sell computer programs. There is a
> big difference between the rights of purchasers (outright buy the code)
> and legitimate acquirers (who merely liscence the code under a set of
> restrictions).

    This is why software companies _claim_ to license computer programs (in 
the mass market - what two companies or people do between themselves is a 
different matter).   Saying something 3 times doesn't make it true.  ProCD is 
an ugly precedent in this matter, but there are older ones, and (to my 
knowledge) the issue hasn't been before the USSC yet.   Of course, this is an 
international list, YMMV.   You can take the opposition to UCITA (and UCC 2B 
before it) as signs that there are serious questions about the legitimacy of 
these claims of licensing rather than transfering ownership.
    It might be useful to have an installer tool that parses installation 
data but discards license ultimatums.  I'd prefer such things were 
unilaterally declared worthless (as they are essentially, in the mass-market 
context, attempts to "privatize" copyright law with state contract law, thus 
should be preempted), so that website agreements could be considered in the 
same class [because there's a difference between purchasing the CD at a store 
that has some nonsense fine print, and downloading something after explicit 
consent that an installer wouldn't address].

Lynn