|
|
|
| Authored by: Anonymous on Monday, May 14 2012 @ 01:26 PM EDT |
(I'm not trolling.)
However, I suspect that the judges and lawyers involved are
not operating from a position of incomplete reasoning and
are not complete idiots.
Law is a collection of rules. Those rules, for a lot of
intractable organizational reasons, are almost always highly
nonoptimal. (If you want optimal rules, please elect me
dictator-for-life. I'm fairly average, and would do better
than Congress.) Experience has shown that it is more
important that laws be predictable and consistent than
correct. That judge feels that consistency in law dictates
that software patents be kept.*
Patent law is obviously highly dysfunctional for software
patents. Probably the biggest problem is simply that, with
so many ridiculous, incomprehensible, probably invalid
patents out there, it is essentially impossible to predict
who or what might sue you. The judge does seem amenable to
changes in the law and its application to software patents.
However, the judge does not seem to be amenable to the
'there's nothing patentable in software because it is math'.
I suspect that the issue is that there's an arguable
societal benefit to patenting some applications of math and
prior precedent already exists.
At this point, legislative rather than judicial changes
probably make more sense. (I'm not sure. Judges often move
faster than Congress and the current mess in software
patents may be best solved by invalidating the whole lot.
Bear in mind that a 20 year patent term is not a lot of time
relative to changes in law.)
Personally, I'd recommend:
(a) A presumption that patents are invalid until validated.
(That, by an overwhelming fraction, is the way to bet.
Validation should involve inspection by a collection of
experts in the field and permit negative briefs by
interested parties.)
(b) That software and business methods have 7 year terms,
starting from date of filing,
(20 years is too long for software. It often takes 20 years
to turn a profit from hardware, so 20 years is reasonable.)
(c) That the standards of non-obviousness be _significantly_
raised
(seriously...glakk...I'm not skilled in the art...and the
oracle-google patents are simple. So are all of Apple's
patents. A patent needs to be non-obvious enough that I'd
save time by looking it up instead of instantly
'reinventing' it.)
(d) That claims be declared invalid unless written so as to
be comprehensible to a layman.
(current patents are essentially useless as disclosure. If
someone proposed patent searching as a way to keep up with
recent developments, they'd get laughter as a response.
That isn't a good thing.)
(e) That unpracticed patents be very, very limited in terms
of damages. (Just a side note...patents should never block
the creation of new, different products.) Basically, unless
the patent owner can establish that they're already doing
the same thing, the infringer should never be liable for
more than a modest percentage of net profits.
The long-term effect would be that the number of 'valid'
patents would shrink by roughly 30-100x and that the
remaining 1-3% of patents would be useful references. This
would largely eliminate the 'patent thicket' problem and,
overall, result in a patent system that better boosted
innovation.
For the physical and life sciences, I believe that these
changes would result in a patent system that strongly
benefited society. For software patents, I'm not certain
that the above would result in a system that was a net
benefit. I'm also not certain that the resulting system
would be a net loss. I could easily believe the argument
that, given the current structure of the software industry,
nothing that involves patents is of any use. However, I
could also buy the notion that, given the probable slowdown
in Moore's law, it is reasonably likely that the software
industry will soon (20-30 years) transition to a
conventional model and this debate will be largely moot. Eh.
Um. Release the hounds?
--Erwin
*Bear in mind that, as ridiculous as the software patents
you've seen are...most other patents in other fields are
similar in ridiculousness to anyone skilled in that art.
I've seen recent patents based on technology developed
before the author was born.
Mind you, I suspect the prior precedents allowing business
method patents are somewhat questionable.
[ Reply to This | Parent | # ]
|
|
|
|
|
