1 2 3 4 5 6 7 8 9
Not to mention that if the builders knew it, then it wouldn't be novel.

But that the builders knew 3-4-5 needn't mean they understood the generalization.

Good point. You'd have to distinguish your notion from the prior art, but that's doable (using wording like "a right triangle of arbitrary size"). You'd probably have a hard time getting past the "law of nature" though. And it's not clear what you'd patent. If it was merely constructing triangles with those dimensions, obviously everybody already does it, whether they understand how or not, so it's not new. So it would have to be something like a "method for calculating the length of a side of a triangle", and I'd expect that to get rejected under the argument that you didn't actually invent anything, you merely discovered what the relationship already is.

On the other hand, if you discovered a novel way of computing the length of the side (say, using a laser and some mirrors), you'd have no problem.
So if it's a general law, you can't take out a patent, but if it's a mathematical method, you can?

That's my understanding. A method seems to have to have a series of steps that don't merely reflect the law.
At this point, I wonder about copyright in prime numbers: as a non-mathematician I'd guess that you might reasonably have copyright

You can't copyright a number. Period, I believe. I'm fairly certain that you can't even trademark one, which is why Intel named the follow-on to the 286, 386, and 486, "Pentium" and why car manufacturers make sure that their model numbers have letters in them as well.
or a patent in some method for discovering prime numbers, but not the number itself. You haven't invented the number, after all; and that it's prime looks like a "law of nature" to me.

Right. (Standard disclaimers.)

Evan Kirshenbaum + HP Laboratories >The law of supply and demand tells us
1501 Page Mill Road, 1U, MS 1141 >that when the price of something isPalo Alto, CA 94304 >artificially set below market level,
... } I don't know what's been happening lately, but it used to be } the only thing you had ... thing they wouldn't even look at? Is that by law or is it right there in the Constitution? (ObBob: IWLIUBICBBJN)

You are correct sir. It's at least in the MPEP and I suspect there's case law that talks about perpetual-motion machines being inherently unpatentable on grounds of total bogosity. I, too, suspect more than a little spark of humor.

Steny '08!
Teachers: We supply a list of EFL job vacancies
So if it's a general law, you can't take out a patent, but if it's a mathematical method, you can?

It still has to be tied in some formalistic way to the real world, like a method implemented on a computer in some way. Plus, the mathematical method can't be of a sort of 'general law' sort (e.g., a mathematical method that computes something in accordance with the Pythagorean formula would be no good (even if it were just discovered)). Outside the US there are generally greater restrictions, but these are currently in flux, from what I understand.

Steny '08!
I don't know what's been happening lately, but it used to be the only thing you had to provide a working model for was a perpetual-motion machine.

In earlier days (up till 1880) a working model was gnerally required. See for more than you probably care to know about the subject.

Ray Heindl
(remove the Xs to reply to: (Email Removed))
At this point, I wonder about copyright in prime numbers: ... method for discovering prime numbers, but not the number itself.

Nothing reasonable about it. It's all just a matter of what a particular judge or court happens to decide.

Copyright has evolved around works of artistic and literary character (irrespective of artistic or literary merit, if you can work that out). A number tout simple falls short.

Paul
In bocca al Lupo!
Students: Are you brave enough to let our tutors analyse your pronunciation?
So if it's a general law, you can't take out a patent, but if it's a mathematical method, you can?

That's my understanding. A method seems to have to have a series of steps that don't merely reflect the law.

A mathematical method is definitely a no-no in Europe.
At this point, I wonder about copyright in prime numbers: as a non-mathematician I'd guess that you might reasonably have copyright

You can't copyright a number. Period, I believe. I'm fairly certain that you can't even trademark one, which is why ... 386, and 486, "Pentium" and why car manufacturers make sure that their model numbers have letters in them as well.

On the other hand, you can have a trademark for a number in Europe. Boeing owns 747, for example.
or a patent in some method for discovering prime numbers, ... it's prime looks like a "law of nature" to me.

Right. (Standard disclaimers.)

See above for the method: a European no-no.

Paul
In bocca al Lupo!
I am told that Unisys Corporation have patents which are ... planning the planting of your vegetable patch, go right ahead.

I don't believe that's generally correct under US law, as I was saying to Erk. The patent monopoly isn't restricted to a particular contemplated use.

It depends, doesn't it?
If you make some software that performs the same data compression algorithm but for vegetable patch planning purposes, that sounds like patent infringement to me. Nay? Of course in a formalistic sense the algorithm isn't monopolized, but we can ignore that.

We can't really say without looking at the source documents. If the patent claims only making compressed data files, I'll stick to my comment; but if it in fact claims something else, all bets are off until we know what it does claim.

Paul
In bocca al Lupo!
When was that? I'm pretty sure that the three main ... or a method for doing something in a novel way).

But I thought the "method" had to be demonstrable in some physical way, ie, one had to have actually built ... drawings of such an implementation complete with scientific data proving (to the satisfaction of the examiner) that it would work.

You don't usually have to prove it would work. You have to describe how to make it work.
You would need to provide data if, say, the prior art says use a metal, and you have discovered that, unexpectedly, titanium gives much better results than known theory predicts.

Paul
In bocca al Lupo!
Students: We have free audio pronunciation exercises.
You can't copyright a number. Period, I believe. I'm fairly ... that their model numbers have letters in them as well.

On the other hand, you can have a trademark for a number in Europe. Boeing owns 747, for example.

I just checked. Also in the USA.

Paul
In bocca al Lupo!
Show more