All you're doing with the cherry picking is showing that the system isn't perfect, not that it's broken.
[1] http://kotaku.com/two-melons-just-sold-for-us-15-730-in-japa...
That's demagogy. Completely or not completely is irrelevant. If it's passing through tons of patents like that and doesn't provide adequate means to prevent patent abuse - it requires a thorough reform. Those who claim that the system is not broken want to keep the status quo, because they benefit from the junk patents and patent abuse.
> That's demagogy.
Uh... what? I think the claim that the system is not completely broken is very fair. If you want to prove it wrong, show that 90%+ of the patents are junk.
> Completely or not completely is irrelevant. If it's passing through tons of patents
It doesn't matter which patents pass, what matters is which absurd patents go to court and win.
So, let's see, how many court cases have we seen this past decade where absolutely absurd patents ended up winning in court?
Don't get me wrong, I think the system has problems, I just don't think it's broken and needs to be completely abolished.
We need to find a way to eradicate patent trolls and there are a lot of potential solutions to this. For example, if company A acquires company B, then company B's patents become void and null unless A can prove these patents are applicable to its business. Blam. All patent trolls vanish.
Once you eliminate the problem of patent trolls, you realize that the idea behind software patents is sound and actually promotes innovation, start ups and entrepreneurship.
Eliminating patent trolls would eliminate that problem for large corporations, although it would still cost them million of dollars to apply for patents solely to cancel out others' patents through cross-licensing. Start ups are even more screwed because all of the large incumbents still each have a huge arsenal of broad software patents and can bury any software start up in litigation who tries to compete with them. This also makes having a software patent essentially useless for a start up because any attempt to assert it against an incumbent would only result in counterclaims and ruinous litigation costs. The idea that a small practicing entity can realistically assert a software patent against a large tech corporation is beyond belief. And that's before you even consider the six-figure cost of paying lawyers to get a patent through the patent office, which then generally isn't even issued until the technology at issue is either obsolete and disused or so common and ingrained in the status quo that issuing an effectively retroactive patent on it without prior notice to those who have since adopted it is inherently unreasonable.
Software patents are fundamentally useless and should be abolished.
Gigantic, established companies will always have an edge, regardless of patents. Abolishing the system and creating a wild-west scenario is ridiculous.
Reform and overhaul using principles guided by rational, current, and quantifiable evaluation, with an established timeline for recurring review and reform. That's what we need (in many more systems than patents as well; education, I'm looking at you).
Life isn't always binary. There are more answers than all or nothing.
"If they don't use that, then they can just use their connections to the government to legislate start-ups out of existence." In a way this is exactly what the current patent system accomplishes. Except it's through litigation, not legislation (but it's legislation allowing for the grounds of litigation).
"If they don't use that, they can just buy the property the start-up physically occupies and burn it to the ground." Now you're just being absurd.
"Gigantic, established companies will always have an edge, regardless of patents. Abolishing the system and creating a wild-west scenario is ridiculous." What we have now is a wild-west scenario where submarine patents and overly broad patents allow trolls (and even practicing entities looking to squash competition) can simply litigate at will and the costs of defending overwhelm their competition (unless they're equally large or play the same game with cross licensing patents).
Fairness has nothing to do with it. It's obvious that large entities have more power than small entities, but that's hardly an excuse to give them more. Especially when the argument being made is that it will help small entities which it empirically does not.
> Gigantic, established companies will always have an edge, regardless of patents. Abolishing the system and creating a wild-west scenario is ridiculous.
There should be corollary to Godwin's law that anyone who describes the tech industry as the wild west automatically loses the argument. There are no gunslingers here. You've already listed plenty of reasons why Microsoft is not going to be felled overnight by some startup in the absence of software patents.
> Reform and overhaul using principles guided by rational, current, and quantifiable evaluation, with an established timeline for recurring review and reform.
There is nothing to reform. The essential failure isn't in the specifics of the law, it's in the concept of patenting what is fundamentally information. Under the same law the patent system works largely as it's expected to in the pharmaceutical industry or the auto industry. It fails spectacularly in the software industry because software is not supposed to be patentable subject matter.
It would be if they hadn't already broken the system by letting so many frivolous patents through.
The original view of patent eligibility from those who encoded the system into the US Constitution was VERY different than that of today.
See, for example:
http://www.temple.edu/lawschool/dpost/mcphersonletter.html
If Jefferson had any idea of what the current patent system would look like and how little it had to do with bringing to bear really novel inventions with high social value (compared to being mostly a rent seeking mechanism used to make a quick buck via iffy, vague claims on all sorts of obvious ideas), I'm absolutely certain he would never have allowed for the creation of the system on his watch.
It would be silly to suppose that they spend the entire time in examination. That's why it should be obvious that I think they have ample time to correct their mistakes.
You do have a point insofar as the entire "game" of patent grants is rigged, such that the examiners are graded on how many patents are processed, while the examinees are allowed to resubmit filings as many times as they like until they pass (subject to fees, of course).
Surely we can agree that "not perfect" is not equivalent to "broken"?
For example, a startup with hardly any funding comes up with a very innovative idea to tackle a problem, they start working on it with very little funding. Facebook grabs the idea, implements it in a month, exposes it to a billion users and the startup just dies.
What's the incentive for the next innovators to create a startup in such conditions?
"What's the point of trying to create something novel since Facebook or Apple will just copy us and crush us".
Not exactly a perfect system now, is it?
That makes an assumption that software patents should be used for preventing others from "grabbing the idea". That's invalid. You can protect your software with copyright, or for example make it open and ensure contributions from derivative works with copyleft. But to be competitive one doesn't need software patents which cause all kind of crooked damage.
They wouldn't have to, it's much cheaper and faster for Facebook to just acquire them.
Make no mistakes: whenever you read in the news that Big Company just acquired Small Company, a lot of this is justified by the patents that Small Company filed. It is much faster and cheaper for Big Company to acquire that company, its patents, its team and its products than spending years in court suing them.
That's bunk. It can happen that companies are acquired for patents. But the above was the case when it's acquired for technology. That's a very different thing.
Which is why software patents are so useless! Big Company needs a team and code, and it gets the product to market faster to buy the team who has already written and tested the code than to pay a new team the same money to write it again. Meanwhile the patent is useless because the cost of ligating a patent claim is an inconvenience for Big Company but a mortal danger for the startup.
However it's clear that banning software patents at least would already improve it. I.e. such workaround is better than the current level of broken.