🚀 go-pugleaf

Let me restrict attention to just one scenario: John Shmoe writes a
program that uses the g++ libraries. He sells the source to the program,
for a hefty $150 per tape. To install the program, the recipients need
only compile it---under g++, of course.

This is the ``GPL Library Scenario.'' Notice that someone could get into
the scenario without ever seeing any g++ source, or even the GPL.

The FSF has repeatedly claimed that such a program must fall under the
GPL, by section 2(b). I find this virus effect an outrageous expression
of a morally corrupt attempt to trap people into accepting a philosophy
that should be a matter of individual choice, but I'll stick to the
legal reasons why the FSF's position is untenable. The GPL has many
other problems, but this is by far the most important, as well as the
most clear-cut.

In article <35894@ucbvax.BERKELEY.EDU> jbuck@galileo.berkeley.edu.UUCP (Joe Buck) writes:
> In article <976:Apr2909:03:3890@stealth.acf.nyu.edu> brnstnd@stealth.acf.nyu.edu (Dan Bernstein) writes:
  [ I point out that shrink-wrap license enforcement has been struck down ]
> With shrink-wrap licensing, you pay money before seeing the license.

No. A shrink-wrap license is (literally) a license that you supposedly
agree to by breaking the plastic---even after you've legally obtained
the product. Whether ``legally'' refers to buying or to something else
is irrelevant. The principle that shrink-wrap licenses are invalid
extends to any license that purports to apply conditions not stated
until after receipt.

Maybe you don't understand the point of licenses. For the distributor of
a work, selling a copy means giving away ownership of that copy; and
then the owner has certain permanent rights, such as the right to keep
the copy forever and the rights under the First Sale doctrine. Licenses
can remove these ``disadvantages'': if the distributor can get the buyer
to agree that ownership of the copy remains with the distributor, then
he can restrict the buyer's rights much more effectively.

Say you walk into ComputerLand, see a program on a shelf, and buy it.
That copy is now yours. You leave the store, open the box, and see a
shrink-wrapped package with a license showing through the shrink-wrap.
The license says, basically, ``We retain ownership of this product. You
don't get ownership rights. If you disobey these terms, you must destroy
all copies. We do allow you to use the product or make one backup copy.
By breaking the plastic you agree to this license.''

Now it's patently obvious that the license is unenforceable: you didn't
agree to those terms when you paid. Software companies tried to get
around this with several arguments, most importantly that since you
could return the unopened package for a full refund, your payment really
occurred the moment you opened the package. Fortunately, common sense
triumphed in Federal court: only your legal ownership of the package
mattered, not the distributors' concept of where money went when. Now
we're seeing things like Borland's quite straightforward ``you can use
this just like a book.''


Why is the GPL unenforceable? If I anonymous ftp to MIT and pick up
Emacs, I don't (can't!) see the license until after I've received the
entire package. But at that point the copy is mine, and I don't have to
pay attention to section 2(b). In the GPL Library Scenario, John is
doing nothing outside his normal ownership rights: he's neither
modifying g++, nor distributing it, nor using it in public. So the FSF's
position here is completely unfounded.

> if you read the license and don't like the terms,
> you're free to delete the code.

That's exactly the argument they struck down in Louisiana. It doesn't
make the license enforceable.

> If you paid for a tape, you've paid
> a fee for distribution (the labor of making the tape, the cost of the
> tape, and the cost of shipping).

And once I've received the tape, it's mine, unless I agreed otherwise
beforehand. Get that straight.

> >If you're an author, and you've been selling a
> >product derived from GNU code without realizing the effects of the GPL,
> >and now the FSF is trying to steal your profits: Read 17 USC!
> 
> Give it up.  If the above applies to you, you're either incredibly stupid
> or a thief.

I've always been very careful to keep GNU stuff separate from the rest
of the system, and I distribute my UNIX software for free anyway, so I
have no personal problems with the GPL. I'm mainly worried about poor
John in the GPL Library Scenario, as well as the morals that let you
think you can call him a thief.

> Every module of every GNU program is clearly labeled.

Bullshit. I can use g++ without seeing even a mutter of the GPL. Nothing
in the g++ installation instructions even gives a hint that there's a
potential problem.

> The only case where a software vendor could be confused is if he doesn't
> realize that some GNU code has been linked in as part of a library,

I'm glad you recognize that the FSF has no rights here, but I don't
think it's John who's confused.

> The GNU Public License doesn't force you to give your code away; you
> have several options:
> 	1) Distribute your code with the GNU code according to the
> 	GNU License.

To some people, that's giving your code away.

> 	2) Develop with Gnu tools, but make sure that everything you
> 	ship is your own.

I'll believe that the FSF understands this when I see it.

> 	3) Use your combined software within your own organization,
> 	but don't distribute it to anyone else.

That's a very bad idea. A disgruntled employee could then force you to
make the source code free, and your whimpered ``sorry, Your Honor, I
didn't understand what distribution meant'' wouldn't go over too well.

---Dan