[Patents] Re: [Patents] Possible Patent Solutions.

[Jean-Paul Smets jp@smets.com]

I like the ideas but I wonder is they can be implemented within the
current law framework

Nick Fleming wrote:
> 1) A company can only take legal action against
>  another for ONE patent, per application that
>  another company owns NO MATTER HOW MANY
>  PATENTS they infringe on.
> 
>  Think about this for a minute, because the
>  benefits to everyone are not obvious (IMHO).
> 
>  (i) For the little guy, this means that
>   IBM can only sue you on a per application
>   basis- not per patent. And then, they
>   should only be able to demand a reasonable
>   royalty - nothing more. (No backpayments
>   either).

OK, but this does not change the fact that if the small guy is sued for
infringement and loses, he/she will be responsible for the supposed
potential lost sales of the big guy which can be a tremendously high
amount. 

However, if the small guy was contacted for contributory infringement,
this risk would disappear because of the way contributory infringement
is treated in the law.

>  (ii) For Giants like IBM, this would mean that
>   their income from patents would actually
>   go UP. Why ? Because of the lower threat
>   to being sued off the face of the earth,
>   more developers would start using their
>   technology hence, more royalties. As a
>   famous supermarket owner once said :
>   "pile it high, sell it cheap"
>   If all I had to do was pay a 10p royalty
>   to IBM for every unit I sold - seems fine
>   to me..

I know this argument. It basicaly says that if the infringer (or
contributory infringer) only takes the risk of having to increase the
price of its software in order to pay one or more patent licenses,
everyone is happy : the patent owner who gets money and the software
developer who can develop with no so much risk.

>  (iii) You can innovate freely, knowing that your
>   legal costs will be minimal.

This is true only if you suppose that the outcome of a patent conflict
is paying a minimal amount of money for patents. But this is not the
case with infringement which may result in having to pay for previous
lost sales due to the infringer.

>  (iv) It scales well. The larger the project, the
>   more chance of infringement BUT only one
>   patent fee, rather than dozens from each
>   company.

The risk is that one company will create one subsidiary per patent.
Incorporating in the delaware is only a few USD. It would be very
profitable to create one subsidiary per patent in order to be able to
grab more licenses.
 
> I can only think of one flaw in this, but it does explain
> why IBM patents everything in sight. If you infringe on a
> number of patents held by a LOT of companies THEN it gets
> expensive. IBM's response to this (being sued by everyone)
> is to patent like mad in order to overcome being sued every
> working day.

OK.
 
> IBM are actually doing us all a long-term favour, strange
> as it seems. By patenting everything, it will all be
> public domain in twenty years time - earlier if they forget
> to renew patent licences.

They would actually d both a long term and short term favour is they
published software without patenting its related processes. Publishing
software creates an anteriority which makes it impossible, in Europe, to
get a valid patent on any process implemented by the published software.

> 2) The second idea is a change to the way the system
>  of paying for patent licences is done. Instead of
>  a variable fee (a reasonable royalty) determined
>  who knows how, the fee should be taken as a
>  FIXED percentage of a businesses PROFITS. (The
>  percentage amount should be fixed when the patent
>  is issued to prevent perceived unfair price rises
>  such as the GIF licencing problems).

Patent law gives a property title to the patent owner. The patent owner
can only be free to decide how he wants and IFhe wants to be paid for
licenses. 

One thing that could be excluded from patent (in Europe) is the
development of open source software by saying that it is similar to a
research process and that patents can not be enforced on research processes.

>  By only targeting profit making companies,
>  free software, the public and non-profit organisations
>  will be able to continue as before. This is an
>  important point, to preserve the publics freedoms
>  and also, to allow the useful development of free
>  software to continue.

But Thomson etc. will never accept to allow people to develop free DVD 
decoders.

Regards,

JPS.

PS. There is one strong common point between your idea and the paper I 
published
        http://www.smets.com/it/policy/useright/useright.pdf
It is the fact that if the only risk of software patents is that user
may have to pay for using patented techniques implemented by software,
everyone is happy : patent owners who make money, software developers
which may at most have to remove their software from the Internet.