Subject: Re: termless copyright and patents
From: Thomas Lord <lord@emf.net>
Date: Tue, 03 Oct 2006 15:37:48 -0700
Tue, 03 Oct 2006 15:37:48 -0700
simo wrote:
> On Tue, 2006-10-03 at 13:18 -0700, Thomas Lord wrote:
>
>   
>> Teachings, products, etc. used to decide the validity of
>> a patent under the "obviousness" criteria.
>>     
>
> Ok, but I don't see why you are talking of prior art for a granted
> patent.
>   

It's very simple.

Given a set of evidence, labeled "prior art", the law defines
a set of innovations which are "obvious" and therefore
should not be granted patent protection and can not infringe
on any patent.

In other words, patent law imagines a black-box process,
called "skilled practitioner", that takes a set of teachings
as input, and produces a set of innovations as output.  


      teachings  -> skilled practitioner -> obvious inventions

In deciding the validity of a patent, law and jurisprudence
places two limits.   It considers only teachings that existed
when the patent was granted.   It considers the skill level
of practitioners at that same time.

Well, we can re-use the courts ability to evaluate what a
skilled practitioner would do.   We don't have to limit
ourselves to teachings from the distant past or skilled
practitioners from the past.   We can think of the law as
defining a function:

      f  ( set_of_teachings, skill_level ) -> obvious_inventions

We can say that a patentee may only convey a
GPL program if they promise not to assert patent claims
against the act of making, using, or selling a GPL program
if such making, using, or selling is not an "obvious invention"
for teachings which include the works conveyed, and
for a contemporary level of skill.

     f (teachings_including_conveyed_works, contemporary_skill)
     -> covenanted_inventions

That is consistent with the "four freedoms" mission of the
GPL.   It's crystal clear (well, as clear as can be) to courts.


> Once a patent is granted you have to go to a court to invalidate it by
> showing prior art the the poor PTO couldn't find before the grant :)
>
> But unless you want to go into a court I don't see why talking about
> prior art can matter.
>   

I'm not talking about prior art per se.   I'm talking about a
covenant expressed *as if* a conveyed work were prior
art.


> It is better in the sense that protection is extended not only to the
> recipients of the first program but also to downstream recipients.
>
>   

Better than what?   The rules I've been describing protect
downstream recipients so I am not sure what you are
comparing to what.

-t




simo wrote:
On Tue, 2006-10-03 at 13:18 -0700, Thomas Lord wrote:

  
Teachings, products, etc. used to decide the validity of
a patent under the "obviousness" criteria.
    

Ok, but I don't see why you are talking of prior art for a granted
patent.
  

It's very simple.

Given a set of evidence, labeled "prior art", the law defines
a set of innovations which are "obvious" and therefore
should not be granted patent protection and can not infringe
on any patent.

In other words, patent law imagines a black-box process,
called "skilled practitioner", that takes a set of teachings
as input, and produces a set of innovations as output.  


      teachings  -> skilled practitioner -> obvious inventions

In deciding the validity of a patent, law and jurisprudence
places two limits.   It considers only teachings that existed
when the patent was granted.   It considers the skill level
of practitioners at that same time.

Well, we can re-use the courts ability to evaluate what a
skilled practitioner would do.   We don't have to limit
ourselves to teachings from the distant past or skilled
practitioners from the past.   We can think of the law as
defining a function:

      f  ( set_of_teachings, skill_level ) -> obvious_inventions

We can say that a patentee may only convey a
GPL program if they promise not to assert patent claims
against the act of making, using, or selling a GPL program
if such making, using, or selling is not an "obvious invention"
for teachings which include the works conveyed, and
for a contemporary level of skill.

     f (teachings_including_conveyed_works, contemporary_skill)
     -> covenanted_inventions

That is consistent with the "four freedoms" mission of the
GPL.   It's crystal clear (well, as clear as can be) to courts.


Once a patent is granted you have to go to a court to invalidate it by
showing prior art the the poor PTO couldn't find before the grant :)

But unless you want to go into a court I don't see why talking about
prior art can matter.
  

I'm not talking about prior art per se.   I'm talking about a
covenant expressed *as if* a conveyed work were prior
art.


It is better in the sense that protection is extended not only to the
recipients of the first program but also to downstream recipients.

  

Better than what?   The rules I've been describing protect
downstream recipients so I am not sure what you are
comparing to what.

-t