Re: lynx-dev LYNX: another article, on M$ and antitrust and conservative judges

[David Woolley]

> 
> I'm *sort of* being devil's advocate on purpose, but sort of not.  If patents
> are there to make people rich off of their intellectual advances, why leave
> software out of the loop?

Patents are there to encourage innovations and to encourage publication, not
to make people rich.  In return for a patent you are required to publish 
the basis  of the invention.

The problem with patents in general is that they favour the big companies
that can afford lawyers.  The problems with software patents are that:

- the lifecycle time of software is much shorter than the lifetime of patents;

- most of the cost of most software is not the original ideas but the detailed
  coding and testing;

- because patents have not be used in the past, people have not documented
  prior art; however, there is really very little true novelty in software -
  most developments are re-implementations of earlier concepts under a new
  marketing name - the real developments are in faster hardware, making old
  ideas viable.

I do actually have some sympathy for the encryption patents, except in as much
as a lot of the real work seems to have been done on academic/government funds.
To the extent that a commercial organisation has self funded the development of
an algorithm, they have put real money into it.

For everything else I think that confidentiality law and the lead time from
idea to market are sufficient protection.  Even for encryption, I think the
patent lifetime for software should only be the two to three years lifetime
of a commercial software product.