Re: Using non-GPL libraries in a GPL program

[David Kastrup]

Tim Smith <reply_in_group@mouse-potato.com> writes:

> In article <8563sspssm.fsf@lola.goethe.zz>, David Kastrup <dak@gnu.org> 
> wrote:
>> > How come when RMS decides to wax forth legalistically, there never
>> > seem to be either cases or statutes that back his position?
>> 
>> Because the FSF is not a litigation company and so the GPL has been
>> written not as a contract, but a license.  As a consequence, it does not
>
> See, that's a perfect example.  A license is a kind of contract.

Uh no, it isn't.  What the software companies purport to call a license
is a kind of contract.

> It's handled in court under contract law.  But in FSF-land, license
> has some kind of different meaning that doesn't correspond to the one
> used by lawyers and the courts.

Uh no.  It corresponds perfectly with what the lawyers and the courts
call a license.  It does not correspond with what typical software
companies call a license.

>From Bouvier's Law Dictionary, Revised 6th Ed (1856) [bouvier]:

  LICENSE, contracts. A right given by some competent authority to do an act, 
  which without such authority would be illegal. The instrument or writing 
  which secures this right, is also called a license. Vide Ayl. Parerg, 353; 
  15 Vin. Ab. 92; Ang. Wat. Co. 61, 85. 
       2. A license is express or implied. An express license is one which in 
  direct terms authorizes the performance of a certain act; as a license to 
  keep a tavern given by public authority. 
       3. An implied license is one which though not expressly given, may be 
  presumed from the acts of the party having a right to give it. The following 
  are examples of such licenses: 1. When a man knocks at another's door, and 
  it is opened, the act of opening the door licenses the former to enter the 
  house for any lawful purpose. See Hob. 62. A servant is, in consequence of 
  his employment, licensed to admit to the house, those who come on his 
  master's business, but only such persons. Selw. N. P. 999; Cro. Eliz. 246. 
  It may, however, be inferred from circumstances that the servant has 
  authority to invite whom he pleases to the house, for lawful purposes. See 2 
  Greenl. Ev. Sec. 427; Entry. 
       4. A license is either a bare authority, without interest, or it is 
  coupled with an interest. 1. A bare license must be executed by the party to 
  whom it is given in person, and cannot be made over or assigned by him to 
  another; and, being without consideration, may be revoked at pleasure, as 
  long as it remains executory; 39 Hen. VI. M. 12, page 7; but when carried 
  into effect, either partially or altogether, it can only be rescinded, if in 
  its nature it will admit of revocation, by placing the other side in the 
  same situation in which he stood before he entered on its execution. 8 East, 
  R. 308; Palm. 71; S. C. Poph. 151; S. C. 2 Roll. Rep. 143, 152. 
       5.-2. When the license is coupled with an interest the authority 
  conferred is not properly a mere permission, but amounts to a grant, which 
  cannot be revoked, and it may then be assigned to a third person. 5 Hen. V., 
  M. 1, page 1; 2 Mod. 317; 7 Bing. 693; 8 East, 309; 5 B. & C. 221; 7 D. & R. 
  783; Crabb on R. P. Sec. 521 to 525; 14 S. & R 267; 4 S. & R. 241; 2 Eq. 
  Cas. Ab. 522. When the license is coupled with an interest, the formalities 
  essential to confer such interest should be observed. Say. R. 3; 6 East, R. 
  602; 8 East, R. 310, note. See 14 S. & R. 267; 4 S. & R. 241; 2 Eq. Cas. Ab. 
  522; 11 Ad. & El. 34, 39; S. C. 39 Eng, C. L. R. 19. 
  
Note that the "license" from software companies in essence says "Ok, you
are entitled to certain uses according to copyright by buying this
medium.  However, you agree to not make use of those rights.  If you
refuse to relinquish your rights, we refuse to let you do some other
things."

That's a contract because the _interest_ of the software company is
monetary.  What is stipulated as license conditions is not interest per
se.

> Same with the concept of "distribution".  Whatever it is the FSF talks
> about when it talks about software distribution seems to have little
> connection to what the copyright law means when it uses that term.
> Remember when the FSF was claiming that giving someone code that can
> be linked with GPLed software is an implicit distribution of the GPLed
> software, even though it is up to the recipient to decide if they want
> to get and link with the GPLed software, or get and link with other,
> non-GPLed software?

It never claimed anything like that.  You are confusing this with
contributory infringement: doing software distribution in a manner that
_requires_ the recipient to perform steps _on behalf_ of the distributor
without which he would not get what he paid for.

> But only if the GPLed and non-GPLed software have different
> interfaces?  If they have the same interface, then it is no longer an
> implicit distribution.

You are making terms up as you go.  When they have the same interface,
linking the GPLed library in stops being a necessary step for the
recipient for getting working software.

Also when it can be convincingly argued that the offering was _not_ for
functional software but, say, for educational purposes (programming
books containing skeleton code or scraps that don't run on their own),
no infringement by the distributor will be actionable.

> I invite you to find any statutory or case law support for that
> interesting theory.  You'll have a long search, 'cause there is none.

Sure.  Because in the stated form, you made this theory up by yourself.
Not the FSF's problem.

> With Terekhov's claims, I can go read the cases and statues he cites,

Yes.

> because he actually makes *legal* arguments.

Actually, he is quite willing to quote stuff which strictly contradicts
his own statements.  While he feels free to make legal arguments, more
often than not they are pretty absurd and don't hold up to what the
courts read into the cases he cites.

> The FSF and RMS make a lot of legal-sounding arguments that appear to
> actually be them stating what they wish the law to be, using terms
> somewhat similar to, not not always the same as, real legal terms.

Their track record in court is quite better than Terekhov's predictions.
Moglen is paid by Columbia University to teach law.  Nobody gives a dime
for Terekhov's legal advice.

There are reasons for that.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum