Comment by jonmc12 | Hacker News Reader

Comment by jonmc12 | Hacker News Reader

Parent Full thread

Really, we should be free to invent. Option B should be that if an inventor can show a clean history of evolving an idea, it should be allowed to freely enter the market.

Ideas evolve over time, and the originate from something. Perhaps if inventors were willing to register their invention documents through a set of steps mapping to something like the Technology Readiness Level (http://en.wikipedia.org/wiki/Technology_readiness_level), then they could be eligible for immunity against patent based litigation.

But lets say this certified patent wants to protect its IP on the market through litigation. In this case, the inventor could first request certification of the invention in question before any further legal action occurred.

If we could create a certification process that is less than $2000 or so (about the cost of a provisional patent including attorney fees), it seems feasible. This system would seek to encourage a regulated invention process that filters new inventions from even being eligible for litigation. This allows for inventor insurance in a sense - keeping a capped cost for protecting certifiably original ideas while encouraging original invention and discouraging the incentive behind patent lawsuits.

1) What stops me from "cleanly inventing" something that is already patented? There's no way to prove that I derived my work from the patent.

2) How do I know which things need to be certified as inventions? Software patents are so vague that even simple applications can infringe hundreds of them. iOS apps are being sued just for having in-app upgrades.

3) Anything that requires more beurocracy or litigation is only going to make things worse because small players can't even afford to make use of the system, especially when they are fighting against dedicated full-time lawyers. The only alternative to solving that underlying problem is eliminating the ways it can be exploited.

1) Keep logs of any and all materials read, communications, etc. Disclose all sources of information involved during the various phases of invention - specifically reviewing hard drives, mobile and other storage. The inventor should take care to keep logs of all research. In original invention, the trail from research to the idea should be evident if it meets the right patterns - I'm not sure exactly what these patterns would be, but I do think there is a definition that is meaningful enough to considerably aid the situation.

2) Eligible inventions would include anything that a customer pays for - as in the exchange of value designates the scope of invention from a patent enforcement perspective. From a patent defense perspective, as long as the invention logs show no use of another patent (when the attacking inventor requests certification at their own expense). This would also focus invention on creation of value instead of creation of legal IP.

3) I disagree, this proposal is an attempt to let people opt out of a broken system by certifying invention through at least one adopted methodology of original design. Again, think of it like buying insurance - every complex industry requires some sort of overhead to deal with subjective enforcement of laws. The capping of costs reduces the average overhead so much vs the current system where anyone who succeeds in the market pays a big, unpredictable legal tax.

"Keep logs of any and all materials read, communications, etc."

Seriously? Who the fuck has time...

I've worked with computer forensics investigators before - you'd be surprised how much data you can get from your harddrive and mobile devices.. If inventors took a bit of care, and good logging / storage software was available, I think the certification process could be as straightforward as applying for a provisional patent.

Can forensics investigators tell whether I came up with an idea in the shower vs hearing about an existing product?

This is exactly the mindset that leads to the patent system we have today. You are thinking of intellectual property as an idea - its not. Meaningful intellectual property is a functioning method or product that delivers value to people (is this all IP? no, but this the scope of IP I am suggesting protecting). It takes, research, communication, testing, experimentation, failure, and repetition to truly build something that works. And yes, forensics will show the hard drive state and web history logs before, and after you were in the shower and then all the way along the long road to you building something that people actually want.

You hear of an existing product? Who cares? If you can invent something similar without referencing related documentation or code during the actual process of invention keep it out of the legal system.

What leads to the patent system we have today is greed and lawyers. If the system can be exploited to legally steal buckets of cash from successful companies, then it will happen. The mindset of the industry seems to be overwhelmingly opposed to this, but that doesn't matter as long as the law permits it.

A new kind of certification for complex and novel inventions will not solve anything, because it's the simple and obvious inventions that everyone depends on that expose us to patent attacks.

Well, I started with the idea of 'We should be free to invent' - thanks for taking the time to poke holes in one proposal towards this end. I tend to agree that it would be difficult to build a practical certification system, at least with any proven system we've seen today.

Our government provides us many freedoms in the constitution, perhaps an amendment is a better proposal. That way the courts can interpret a more meaningful definition of Freedom to Invent as a way of shaping the current patent system.

I love the idea of a "show your work" patent system, but could it possibly work in real life? Does anyone do anything like this currently?

Scientists do something similar, in that they have to log absolutely everything they do and publish the important parts. Biologists and psychologists are even more on the hook, since they have to demonstrate compliance with experimental ethics regulations. It's a bureaucratic hassle, but it works pretty well.

While you're right that there are some practical issues with this sort of system, one should also recognize that the patent system doesn't current scale well either. You have to check the claims against every publication in existence. Since this isn't possible, this ultimately has to be fought out in court -- yet a granted patent still carries with it a presumption that the patent rights should be upheld.

One could, however, have a system where if two separate teams file patents before either are published, they should not have to license it.

It actually doesn't work, since the only thing required for a patent is the ability for it to be executed. The patent applicant doesn't need to be the one doing it or need to be working with it. All the applicant needs is the knowledge that an expert in the industry believes it can be implemented.

Reply on news.ycombinator.com