🚀 go-pugleaf

In article <PCG.91Aug22184726@aberdb.aber.ac.uk> pcg@aber.ac.uk (Piercarlo Grandi) writes:
>On 19 Aug 91 02:40:43 GMT, kap1@icarus.spc.uchicago.edu (Dietrich Kappe) said:
>
>kap1> In article <PCG.91Aug18210217@aberda.aber.ac.uk> pcg@aber.ac.uk
>kap1> (Piercarlo Grandi) writes:
>
>pcg> The royalty would be by default something like 3-5% (which is
>pcg> little for a new product, where profit margins can be supposed to
>pcg> be something like 30-50%, and a lot for mature products, where they
>pcg> can be supposed to be around 5-10%), and the inventor could reduce
>pcg> (but not increase) it at will.
>
>kap1> One problem I see with it, though, is the cummulative effect of
>kap1> multiple royalties. Suppose a programmer incorporates 10 patented
>kap1> methods in his/her program. The royalties would amount to 30-50%.
>
>Precisely. That's not a problem, at least not a legal problem; legall;y
>the US Constitution only requires that inventions be rewarded by
>granting exclusive economic rights to the invention.
>
>The problem is *economic*: note that I wrote that the 3-5% would be the
>default *and* the maximum, but there would be no minimum. The idea is
>that paying 3-5% of your sales gives you the right to use the patented
>invention without any further ado; if you think that 3-5% is too much,
>you can always negotiate with the inventor for a reduction.
>

I think one *major* difference in standpoints in this thread is a
matter of degree...

Normal *material* inventions or manufactured goods *might* contain 10
parts that have been patented by others. Normal *software* inventions
or production systems generally contain hundreds if not thousands of 
patented algorithms. A maximum 3-5% royalty might be fine for material
inventions, but its absurd for software. In the future, if the current
trend in software patents continue, there won't be hundreds or
thousands of patented algorithms in a program, there will be MILLIONS
then your 3-5% royalties would be a magnificent anchor! Think of all
the man hours spent just trying to get agreements to use a
patented algorithm for whatever royalty...lets see, say 10,000
patented algorithms in a software package. To get an agreement, say 
at least 2 letters written, a contract filled out & signed, get a
lawyer to approve the contract, maybe a few telephone calls, let say
two man hours spent per patented algorithm...hmmm 2 man hours per
algorithm times 10,000 algorithm equals 20,000 man hours...lets see...
40 hours a week times 50 weeks a year (don't forget holidays) comes to
2000 man hours per man year...hmmmm...that means you need to employ
10 people for 1 year *just* to handle the license agreement. The
salaries of the people needed would be higher than minimum wage, lets
say $15,000 per year, with benifits and administration added in, would
come to around $25,000 per year per person (very ball park figures).
Lets see...this company would have to spend $250,000 *just* to secure
all the licensing agreements to use the patented algorithms in the 
package. and if this package is going to be distributed for free???
Not bad...10 people hired to push paper for a year, and then the
product brings in NO revenue??? guess the company will have to go into
some other business, or start manufacturing money...

sounds almost like the music industry..."money for nothing..."

John