And I agree with him, that neither of these patented "inventions" is novel or
nonobvious. Its ridiculous that standard CS techniques dating back at least to
the 60's, can end up being patented in the 90's.
The worst part is that
NEARLY ALL SOFTWARE PATENTS ARE THIS BAD.
A very, very small percentage
of them describe something that is actually clever enough that most programmers,
placed in the same situation, would never have thought of it themselves. But
the vast majority of software patents just cover the everyday work product of
programmers. If you took ten programmers and gave them the same problem to
solve, there's usually a good chance that one or more of them would come up with
the patented idea as their solution (especially if its the best _and most
obvious_ idea for solving that particular problem).
Why is this, you ask?
Simple. In most companies, the way software patents come into existence works
like this. Programmers solve problems (maybe even with clever solutions) and
build software, and the company sells it. And then the lawyers come around to
the programmers and say, "Did you invent anything lately that we can patent?
You get a $500 bonus for each idea we get a patent on." The programmer then
reviews his last two months of work and pulls out five or ten of the most clever
things he thought up, and immediately discards most of them because they are so
obvious that he can't pretend even to himself that they deserve patent
protection. But the last one or two he's not so sure about, so he describes it
to the lawyer in a couple of sentences. And the lawyer's eyes light up, and
thus begins several back-and-forth sessions where the programmer clearly
explains things and draws diagrams, and the lawyer translates everything the
programmer wrote into unreadable legal gobbledygook.
Seriously! I've seen
this process first-hand at multiple past employers. (Fortunately, at my current
employer everybody is so busy trying to meet deadlines that nobody has time to
worry about stupid things like applying for patents.)
When the application
is dozens of pages long and not even the programmer understands what it talks
about anymore, they attach a bunch of prior art and ship it off to the patent
office along with the fee (a few thousand dollars) and wait for the first
rejection. (Hopefully the programmer is no longer involved by this point.)
When it comes back, they tweak it to get around any objections of the patent
office, maybe by limiting claims or maybe just by changing the terminology so
that it purports to not claim obviously-unpatentable things anymore (even though
it's still a SOFTWARE PATENT which is a ridiculous thing that should never have
been allowed to exist in the first place, oh well).
Eventually, the company
gets their patent, the programmer gets his $500 bonus, and for the next 17
years, everyone else who uses even vaguely similar ideas in their software is in
danger of being sued for infringement and having to spend millions of dollars to
defend themselves.
Can anyone see what's wrong with this
picture?
The programmer gets a $500 incentive, and it only costs him
a few hundred dollars worth of his time. The lawyers make a few hundred dollars
for their hours. The company spends a few thousand dollars each time they apply
for the patent. But someday, they can use it to sue some other productive
member of society for millions of dollars in damages! Unfortunately, those
millions of dollars don't come from thin air. The very existence of patent
lawsuits (and patent lawyers!) is a significant harm to society, and we would do
well as a society to just do away with them both. [ Reply to This | Parent | # ]
|