Hacker News Reader

Comment by jowiar | Hacker News Reader

> In fact, you can break this law without even knowing that someone else thought of the idea first.

This is the biggest flaw with the current patent system. The reality is if an invention falls into this category, it should not be patentable. Raising the barrier of "obviousness" to a level requiring an invention be transformative to the point that, within the term of the patent, the likelihood that anyone else independently invents the same thing is sufficiently small would go a long way to addressing these problems.

View on HN

>within the term of the patent, the likelihood that anyone else independently invents the same thing is sufficiently small //

The problem then is that there is little to no incentive for the applicant to share the idea. If it's something no one will come up with, then you can keep it secret and benefit from an indefinitely extended monopoly.

The quid pro quo of patents is early information about new inventions in exchange for limited time monopoly. Here you're encouraging inventors/companies to prefer industrial secrets.

If inventions meet your requirements but are kept secret then the public domain will be worse off as no one gets the benefit of being able to privately use/research the invention unless it can be replicated. With a patent the full details are disclosed.

Obviousness is very difficult to assess because some things that took many years developing are so ingenious that they seem obvious. The applicant always has the argument - if it were so obvious how come, in such a well worked field, with such demand for this invention, how come it wasn't already developed? The answer of course is that it wasn't "obvious" [to the skilled practitioner with knowledge of the prior art in the relevant domain].

Ok, but what percentage of software patents are going to be useful in 20 years when they go into the public domain anyway? Doesn't seem like there's a lot of society downside to just keeping them secret vs open but unable to use until 2032.

We don't have software patents (per se) [!!] in Europe.

But anyway ... if it's not useful it doesn't matter that it's patented and if it is then you've got full disclosure. So I'm not sure the percentage is really important. Perhaps the fees should increase at a greater rate to encourage early release of patented inventions that aren't high worth.

Personally I'd limit the term to about 8 years for all patents and dispense with US business method and pure software patents.

It seems odd that the patent system does not take into account whether an invention could be nearly as profitable as a trade secret or on the other hand, could not possibly be kept secret while used for profit.

Any good examples?

I'm having trouble thinking of something that:

1) You can benefit more from as an industrial secret than as a patented invention.

2) Society is harmed by the lack of public disclosure.

Imagine a manufacturing process, and a transformative invention that makes the product 10 or 100 times cheaper to make than any method that does not use the invention. The invention is used in-house, and there is no advantage to be gained from licencing the invention. Distributing the process to be closer to raw materials or customers conveys no advantage over centralized production. Given such a situation, the company with the invention is likely to bet they can keep it secret longer than the term of a patent.

Of course I cannot give any specific examples. They are currently secrets, or they are secrets that died with the company or died with the obsolescence of the product.

Reply on news.ycombinator.com