[Patents] patent pools enforcing per-copy-licensing mode of software development

[PILCH Hartmut phm@a2e.de]

Prof Endres wrote:

> 1. I had pointed to the MP3 example in our November 8 discussion in
> order to explain a few points. Here you have a set of valid patents
> issued between 1986 and 1996 for highly respected and publizised 
> inventions. If you would have looked at their licensing scheme, as
> I suggested, you would have found four options. Three of them provide 
> for a one-time paid-up license fee. If you use those, you can sell 
> your software how you want. You may even give it away for free.

An MP3 encoder?
This is what the license says:

    In general, our licensing model for software mp3 encoder products is
    based on per-unit royalties.

And this general condition is maintained without exception througout the
license.  I quote the full text below so everyone can check.

Even if a one-time paid-up license was possible, this wouldn't change the
fact that open-source distribution becomes effectively impossible, unless
conducted by a state or some other public agent.  If we are to rely on a
public agent, then we might as well impose an obligatory license or change
the patent system.

OK, the licensing conditions is as follows.  If I have overlooked
anything, please point out.  

     _________________________________________________________________
                                      
                           mp3 Software Encoders
                                      
           (patents and object code developed by Fraunhofer IIS)
                                      
         Fraunhofer IIS developed fast, efficient and high-quality
    implementations for mp3 encoding, supporting bitrates from 8 kbps to
        320 kbps, samplerates from 8 kHz to 48 kHz, mono and stereo.
     Evaluation copies are available upon request, after signing an mp3
            evaluation agreement. Please contact us for details.
     * US$ 5.00 per unit
     * US$ 15,000 annual minimum, payable upon signature and each
       following year in January, fully creditable against annual sales.
     _________________________________________________________________
                                      
                           mp3 Software Encoders
                                      
                               (patents-only)
                                      
   If you have developed your own implementation of an mp3 encoder or if
   you have licensed such an implementation from a third party, you need
                           a patent-only license.
     * US$ 2.50 per unit
     * US$ 15,000 annual minimum, payable upon signature and each
       following year in January, fully creditable against annual sales.
     _________________________________________________________________
                                      
                     Limited Demo mp3 Software Encoders
                                      
    In general, our licensing model for software mp3 encoder products is
   based on per-unit royalties. The license agreement includes the right
    to distribute limited demo mp3 encoders free-of-charge to end-users
      for the sole purpose of generating sales of upgrades to full mp3
          encoders, bearing a per-unit royalty as mentioned above.
   The patent royalties for the right to distribute such limited demo mp3
   encoders are (In case you wish to use the Fraunhofer IIS mp3 encoder,
                            the numbers double):
     * Limited to either 20 encodes or 30 days
       No additional payment
       
     * Limited to 56 kpbs, 24 kHz, mono/stereo
       One-time payment of US$ 100,000
       
     * Limited to 96 kpbs, all sample rates, mono/stereo
       One-time payment of US$ 1,000,000
---------------------------------------------------------------

This evidently makes it impossible to produce an opensource or even
shareware MP3 encoder.

> 2. If users of a patent can organize themselves into bargaining groups
> they may be able to negotiate even better conditions than the ones
> offered to everybody else. I had this in mind, when I referred to the
> Open Source movement collectively. 

If such a bargaining group was successful, it would not benefit itself but
all of society equally.  For each individual, the best policy would be not
to participate in the bargaining group but only enjoy its success.  That
is why such a group can only be a public agent or a state.  There is no
"opensource movement" that can collect taxes and bargain on behalf of the
public interest which it represents.  There is only a large number
individual developpers who are willing to cooperate for the public good,
as long as no undue burden is imposed on them.

> I know that you first would have to convince yourself that the other
> party owns something valuable, however.

The question whether the patented "inventions" have any real merit is
irrelevant for this argument.

> 3. Patents are publications. Neither the Bauer/Samelson push-down
> stack patent nor the RSA patent (to which I also referred on Nov. 8)
> had any negative effect on the teaching and the popularity of both
> concepts during the last two to three decades in computer science.  
> Being an expert on patent law, you should know that there are no
> restrictions on the use of patents for either teaching or research.

That is what I said:  from a university perspective you are not directly
touched by patents, because you are only interested in teaching algorithms
and not in pushing forward the communication process of software
development.  

> 4. I have been exposed to the argument before that all useful research 
> (and software development as well) is funded by tax money. It simply is 
> not true. I heard that the portion may be around 7-8% of all research,
> but cannot defend that number. For software product development it may 
> be even lower, depending on how you measure.

A lot of the things that proved fruitful have come from universities or
from the university environment.  Most programming languages, much of
Unix, TeX, WWW, ...

However you may quantify the importance, you will easily arrive at some
figures like 70-80%. 

> 5. That information and knowledge should be treated as public goods is
> part of an extreme political ideology (or maybe an utopia) that I do
> not adhere to.

Your knowledge is private.
You can own it by keeping it secret.
Once you publish your knowledge, it becomes information and therebey a
public good.
There is no way you can own information.
Intellectual Expropriation is an extreme political ideology and definitely
even a utopia that tries to impose something on nature by the force of
regulation.
I say Intellectual Expropriation, because by the artificial ownership of
information, people are actually expropriated of their knowledge, or at
least of their right to properly communicate and use that knowledge.
At least in the case of free/opensource software, use and communication
are not separable.
 
> I agree, however, that no property is without 
> limits. I, for my part, do not want to live in a society where 
> everything I produce is taken away from me. Whether it is potatoes or 
> information is the same in that respect. 

Me too.
That is why I am against Intellectual Expropriation.

> Nor do I want that others are forced to pay for whatever I decide to
> produce. Paying 30-50% tax on one's income to keep our public sector
> alive, is quite a burden already.  By the way, information and
> knowledge are more likely the work product of individuals than are
> food, land or minerals. To me it makes more sense if those are
> socialized than information and knowledge.

You have introduced another parameter that determines the applicability of
a "property" concept here:  whether something is found (discovered) or
created (invented).  What is created is more suitable for property than
what is found.  Moreover physical objects are more suitable than
informational / logical objects.  An algorithm that is "created"
independently in exactly the same way by many people can be viewed as a
discovery.  Beethoven's Ninth Symphony would hardly be "found" by two
authors.  It carries traits of individual creation.  An algorithm doesn't.  
And it is there eternally, independent of our human world. Like Mr.
Pfeiffer said in our podium discussion: it is "like a gold mine.  the
first to arrive there gets the patent and the others may be pitiable, but
such is life".  Which is of course wrong, because Mr. Pfeiffer neglected
the other parameter:  the fact that information, unlike a gold mine is
independent of time and space, simultaneously usable by infinitely many
people.

> As you certainly know, some economies are still trying to recover from
> such experiments.

And some economies are currently being plunged into the troubles of
another ideology based on expropriation.  More accurately, on the
misplacement of property concepts against natural law.

I also call it "real existing proprietarism".
 
> 6. What you have to say about universities and the German Informatics 
> society (GI) is simply derogoratory and unfounded. I do not comment on 
> that.

I might add something else for everybody's information here.

Prof. Endres is a fervent anti-opensource advocate.  In a journal article
recently published in the GI monthly, he says that society has to choose
between true commercial software engineering and the work of hobbyists.  
True software engineering, he says, needs patents, and the hobbyists
should be sacrificed for that.

In the public discussion I had with Prof Endres at Systems in Munich, he
expressed himself in the same way, saying that we must choose to give
preference to the great achievement of formulating an algorithm over the
trivial mundane work of coding.

Another white-haired professor who was present, Prof. Beier, who patented
a B-Tree algorithm (Greg Aharonian got to look at the patent in Berlin and
told me he suspected it to be invalid), said that "cloning" other people's
programs is an immoral practise that needs to be stopped by introducing
patents.

Apparently, these professors feel that their great achievements are not
being valued enough and feel envy toward the "hype" that is gathering
around some young colleagues who actually write great programs.

The introduction of patents apparently is their moment of revenge, beside
being an easy source of revenue for those who don't write programs
and are shielded from infringement lawsuits by the freedom of education
and research.  
 
It has been known for 100 years that patents have very little to do with
any real advances of science.  Traditionally, patents are not associated
with universities but with industrial businesses.  They usually "protect"
some petty little feature that is at best a result of the application of
science.  In software even more so.  

So why this talk "software enineering needs patents" and "we must protect 
the real scientific advances and not mere coding or cloning"?

I can see only an attempt at deception of the public, fuelled by greed,
envy and a desire of revenge.

These basic instincts are so strong, that these professors actually hurry
to speak in the name of the Society for Informatics (GI), whose member
base is largely against software patents and on the whole not sufficiently
informed, since the GI has neglected the subject for years.

> When we first met, I thought there might be some facts that you had not 
> come across to. In the meantime, I have more and more the impression 
> that you prefer to ignore most of them. This makes it difficult to argue 
> with you.

Look into the mirror, as Greg has always told me.
In fact, the weaknesses you see in others first often reflect your own
weaknesses.

Happy New Year!

-phm