Subject: Re: the walls have ears
From: "William C. Cheng" <william@cs.umd.edu>
Date: Thu, 27 May 1999 16:39:22 -0400

  | >After reading this a few times, it seems to me that Richar'd ``measure
  | >of freedom'' is about:
  | >
  | >    freedom for code
  | >
  | >and most people's measure of freedom is about:
  | >
  | >    freedom for people
  | 
  | No, you read what he wrote wrong.  Please see my post.  What you are
  | saying is exactly what GPL opponents have been saying for something
  | like ten years, and it has never been correct, even though *some*
  | GPL opponents (*not* RMS) often use language that suggests they *are*
  | focusing on freedom of code (usually because they write poorly,
  | getting carried away with an abstraction, as that's easy enough to do).
  | 
  | RMS, the FSF, the GPL, and myself, care not one whit about "freedom
  | for code".

Okay...

  | For my part, that's simply because code is not people, it
  | has no rights, it has no responsibilities (the two go hand-in-hand,

My point exactly.  It would be silly to talk about freedom for code!

  | which is why I find the phrase "animal rights" to be nearly 100%
  | absurd), and, to the extent anyone argues (from a viewpoint I happen
  | to agree with) that "ideas are eternal, therefore they cannot be
  | restricted", then nothing we do can actually limit the freedom of *code*.
  | 
  | >Certainly, GPL gives the most freedome for code.  But it clearly gives
  | >the people receiving the code less choices.  Can't we focus on the
  | >people when we talk about freedom?  After all, a license is passed from
  | >one person to another and not by code having babies.
  | 
  | GPL does not give the "most freedom for code" even if you look at that
  | as the point of it, because the code cannot decide for itself to go
  | into a distributed proprietary product.  Public-domain code gives
  | the most freedom, from that point of view.  That code can get up in
  | the morning, find itself a nice, warm spot in MS Windows 2000, cozy
  | up next to some BSD code, and ride on out to public release, without
  | having to ask for approval from anyone.

I've heard many times from the proponents of GPL that, under GPL, the
code would be more ``free'' than other licenses even though other
licenses give the recipiens more freedom to choose.  And I agreed with
them, although the notion of code freedom is silly.

  | Indeed, let's focus on the freedom of the *people*, not on creating
  | tired, old, straw-man arguments.

Okay...

  | Are you free to distribute the *source code* to the products you, your
  | company, your neighbors, and your friends are using *right now*?
  | 
  | The answer is "yes" if that code contains GPL'ed code, quite likely
  | "no" otherwise.
  | 
  | That's not the *complete* definition of freedom, but it's the viewpoint RMS
  | was getting across.

But that's not good enough!  What about the freedom *not* to distribute
source code?  Isn't that also freedom?  Freedom for people is about
choices, right?

Are you free to sell your product without making your source code
availabel to your customer?  Clearly, you can't if your code contains
GPL'ed code!  So, using GPL'ed code limits my freedom as opposed to
using some other free liecneses.  How is it possible to argue that
I received more freedom when I chose to receive GPL'ed code instead
of code under another free license?

  | And, you'll note, it has *nothing* to do with "freedom of code", because
  | the *source code* doesn't distribute itself.
  | 
  | (Yet.  :)

You mean one day Microsoft would ask a piece of code, "Where do you want
to go today?"
--
Bill Cheng // bill.cheng@acm.org <URL:http://bourbon.cs.umd.edu:8001/william/>