Re: Intermittent Wiper Inventor Dies

Yup. Another feature I find less obscure but amusing is the idea that a patent is rearward looking. Because you are granted a patent doesn't give you the right to the device, it just means no one else can have that right. I suppose that's to make clear that a patent can always be challenged like if the stuff was hiding in the public domain somewhere. (I'm not suggesting I understand patents, the one Nobody listed and the one I just described are IMO two notable issues.)

No, that's 'novelty,' another requirement.

'Obvious' is subjective, obviously, but there are a few guiding principles. One is that if an idea is so obvious and has commercial value, surely someone would be exploiting it.

Obviousness is, though, not as clear-cut as novelty (being the first to describe something new).

Cheers, James Arthur

It's *supposed* to be another requirement, but the USPTO has tried to scrap the non-obviousness requirement by interpreting it as being equivalent to the novelty requirement.

Given the sum total of existing human knowledge, there are a near-infinite number of "obvious" corollaries. OTOH, there is only a finite amount of human capital available to actually exploit them (or even write them down) right now.

Here is an example of how to get patents on something obvious for which there is no prior art.

First, make a list of computing applications or technologies which involve the use of a communications channel (e.g. email). Then, wait for a new communication technology to be invented, and start submitting patents for "method and apparatus to (e.g.) send email via ".

It seems to have worked well enough for the previous round of "method and apparatus to using a computer network" patents.

You don't need to invent either component; you just need to "invent" the combination of the two.

Constructing a working prototype is easiest if the nature of the components is such that you already have an abstraction layer (e.g. the OS' networking subsystem) to ensure that new "instances" of one class (e.g. communication technologies) will automatically work with all existing instances of the other class (e.g. applications).

You made a definite statement: "The USPTO's preferred definition of non-obvious is...", then described their definition of 'novelty.' I was responding to that.

That's not their definition of 'non-obvious.' The USPTO's preferred definition of "non-obvious" is given in Vol. 35, section 103 of The United States Code (35 USC 103):

and clarified at length by the Manual of Patent Examining Procedure (MPEP) (the governing manual used by examiners), sections 2142-2146:

Those sections (easy to read--they're not difficult or lengthy) describe and clarify official practice. It's not fool-proof, but it's a lot more robust--and there's a lot more to it--than simply 'no pre- existing record.'

(The 'novelty' requirement is described by 35 USC 102 and MPEP 2132.

Right. Everything's been invented, everything's obvious, yet people who could benefit from an 'obvious' variation ignore it because they don't have time to write it down.

Without commenting on your example, an unforeseen combination of two existing things / gadgets / elements can be revolutionary. Some of the best inventions arise from combining seemingly incompatible or unrelated technologies / devices / components.

Best regards, James Arthur

That's not the USPTO's definition, it's the legal definition.

The problem is that the USPTO's interpretation of "non-obvious" is so narrow that even the slightest difference to prior art is deemed non-obvious. IOW, the "if it was so obvious, why isn't it written down" argument.

When USPTO decisions are overturned by the courts, it's invariably a case of them granting bogus patents rather than rejecting valid ones.

No. There's a difference between a near-infinite (or even infinite) number of things and "everything".

No-one has an infinite amount of time.

People won't normally think of a solution until they need one. At the point that they have a problem to solve, they will think of a solution. Often the solution is blindingly obvious, but that doesn't mean that someone else hasn't been granted a patent on it.

I've probably never needed to multiply 69112 by 32536, and I doubt that anyone has published that specific calculation. That doesn't mean that the answer is "non-obvious", or that the first person who computes the answer should get to own it.

That isn't the issue. The issue is that entirely foreseeable combinations get deemed "non-obvious" and patented.

The patent system is currently being implemented as a form of "land rush". Patents are being granted without any regard to the original goal of "promoting the progress of science and useful arts".

In many areas (particularly software), patents are actively hindering progress by making it risky to develop anything without having a team of lawyers looking over your shoulder to tell you when you've stumbled across something that's already patented.

and

Sorry, that IS the USPTO's definition, and from the governing documents.

If you want to argue against a certain silly patent, fine. They exist. Feel free--I'd likely agree.

Examiners can't possibly know everything that is or ever was, so mistakes are inevitable. There are even mechanisms for challenging and voiding such patents.

That, however, is not ALL patents.

That's not their logic. I gave you links to their documented, enforced examination procedure.

Sigh. Unless you had a novel method or machine for doing it, that wouldn't be patentable. And if you did, it should be.

Quantities are not patentable. Integers are not patentable. Algorithms are not patentable (though the machines that embody them can be).

You have strong feelings, but you don't know much about patents.

With best regards, James Arthur

James Arthur snipped-for-privacy@yahoo.com posted to sci.electronics.design:

Gosh, what part of the difference between 35 USC 103 legislated law and some CRF citation of regulation don't you understand?

I don't understand your point. If you've ever prosecuted an application you surely know that the examiner nearly always cites 35 USC 102 or 35 USC 103, objections or rejections which must then be overcome by the applicant.

Sir Mr. Nobody argues 35 USC 103 is not the standard for obviousness, but it clearly is. That, and as further described in the MPEP (which I linked), is what governs their daily practice.

I'm not sure what you mean by "CRF" though...maybe you intended "Code of Federal Regulations"?

Best wishes, James Arthur

James Arthur snipped-for-privacy@yahoo.com posted to sci.electronics.design:

The MPEP is in the same class as CFR (CRF was a typo). It largely repeats and extends 35 USC 103 and 103. To any any extent that fails to conform to the text of _any_ relevant USC text, it is easily challenged in court.