On May 30, 2007, at 11:57 PM, Samantha Atkins wrote:
J. Andrew Rogers wrote:
All patents are ideas and algorithms.
Not quite. Would you patent the quadratic equation? How about
Newtonian approximation? Means of computing logarithms?
The nice thing about algorithms is that there are so many of them.
One of the disingenuous arguments against algorithm patents is that
it prevents people from doing things. That is patently false. All
algorithm patents do is, at worst, make you use an older and less
efficient algorithm to accomplish the exact same thing or expend the
effort to invent your own version. If any average person can churn
out fantastic new algorithms with only nominal effort then it means
that virtually everyone in the software industry is a bloody imbecile
because virtually no one does it. Yes, there are a lot of frivolous
patents (of all types), but there are also non-frivolous ones (of all
types) -- a separate issue.
Patents are meant to spread the fruit of innovation while
encouraging more innovation. Software patents quite arguably fail
at both.
Nothing in this assertion is not equally applicable to *all*
patents. Most non-software patents are frivolous, so an argument on
that basis would be pretty irrelevant. As I originally stated, I'm
looking for consistency and nothing more. Any defense of non-
software patents is equally applicable to software patents, potential
ignorance of that fact notwithstanding. As I already stated, no one
questions chemical process patents, which have been with us for
centuries, yet they map one-to-one onto software patents in every
respect that matters.
A better question is this: what new applications are magically
enabled by a new algorithm, and if the effort is so trivial why
haven't you developed these algorithms? This is where I find the
arguments against software patents to just ooze disingenuous. On one
hand you can solve world hunger and make white's whiter with a
brilliant new algorithm that no one can live without, but at the same
time no one has ever solved it before *and* solving it was a trivial
exercise. Oh yes, and you want to be the person to solve world
hunger using that algorithm, never mind that you could not be
bothered to actually solve such a trivial problem. If you want to
dispense with all patents I am okay with that, but the transparently
contorted rationalizations that people in the software industry like
to use to justify their commercial use of someone else's R&D is
grotesque. Everyone seems to like patents *except* in industries
that affect them in some way, as though that situation is unique.
Do you really want to go back to trade secrets, *particularly* since
the frameworks are in place today on many different levels to make
that stick much better than they used to? This does not bother me
personally, but I don't know that most people have thought it through.
Really? When I would have to consult a considerable patent
database to build most any middleware system that I normally would
simply start designing and implementing on top of some commonly
available parts? In what way would my work and productivity and
creativity be improved by this overhead much less negotiating
licenses for each and every piece that migh be useful or that, more
commonly, was broadly enough claimed in the patent to cover huge
families of possible solutions to similar problems?
The rest of the engineering world works the same way. They just have
a more organized way of dealing with it.
In reality, most applications never come near legitimate patent
disputes. Most people are building boring applications with boring
requirements, and when they do need something special they talk to
the people that invested the R&D to solve the problem in the first
place.
Cheers,
J. Andrew Rogers
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&user_secret=e9e40a7e
