|
|
|
| Authored by: Anonymous on Sunday, October 14 2012 @ 10:48 AM EDT |
I don't disagree with you on any of your points about
software engineering. (Putting aside the recipe/contract off
topic). But I do reach a different conclusion.
Yes, many software engineers might be able to come up with
the same conceptual solution to a problem, but the execution
of this will be harder. A lot of what makes software useful
is the accuracy, time complexity, and all the 'engineering'
parts of software engineering. There is still as much
engineering that goes into software as any other engineering
discipline; software might actually be the most complicated
machines we've ever made.
Yes, a 20 year term for software is far too long, as the
speed of software evolution far outstrips the speed of other
industries. I'd suggest an exception for software: 10 year
maximum term, probably even as low as 2. However, I'd also
posit that 20 years is too long for most industries
considering the accelerating speed of innovation that we've
seen since patent laws were originally drafted.
Yes, I agree that software patents at the moment are so
broad that they effectively prevent other people from
solving the same problem. This shouldn't be allowed for any
kind of patent.
Yes, I agree that software patents at the moment are so
broad that they don't really give any instruction on how the
solution is implemented. However, its a fine line: you say
that millions of people will invent the solution without
knowing about the patent - so obviously only a very thin
guideline is needed for people to be able to implement it at
all.
Yes, I agree that a lot of software engineering is realising
the problem, and the solution becomes obvious. Can that
still be patented innovation? Henry Ford and his 'faster
horse'.
However, I don't think any of these particular points means
that software should not be patentable subject matter. It
sometimes requires great innovative and engineering effort
to come up with software. Some software inventions have been
possibly the most influential, useful and world-changing
inventions ever. To deny that kind of innovation the right
to be protected via patent seems ridiculous.
There just needs to be a much broader view of 'obviousness'
and 'prior art', and a much stricter view of describing the
methodology of the patent.
In effect there are four parts here:
1) Should software be patentable subject matter? I say yes,
for the reasons above.
2) Ethically, should software be patented? Some people think
yes, some no.
3) Does software meet the other criteria for patentability?
I say most definitely: software can be novel, useful and
non-obvious.
4) Can software be effectively patented? This is the most
difficult part in my opinion. The specification must be
specific to patent the actual solution, rather than the
problem. But it must be abstract enough that it still grants
some protection - get too specific in the details and it
becomes trivially easy to work around.
It's a difficult problem. And it is being made more
difficult by the fact that the law is being based upon
things that are nothing at all like software. We're seeing
the same problems in pretty much everything digital as law-
makers scramble to make metaphors. Is the internet like
radio, or tv, or post-it-notes, or... . Fundamentally, I
think that software (and other digital phenomena) need to be
put in their own classes, and quickly, before the US law
situation wreaks more havoc on the rest of the world.
[ Reply to This | Parent | # ]
|
|
|
|
|
