And how is novelty to be measured?

Every invention relies on one, or often many, prior inventions or discoveries. Patent applications already contain a list of such "prior art". I suggest that the term of a patent should be equal to the number of days between the latest prior art invention and the one being patented.

When all the precursors are in place (other inventions, scientific discoveries, market conditions, etc.), the new thing "can be invented". The number of days that something *could have been* invented but wasn't, is used as a proxy for how novel it is.

The effect we are looking for is that obvious things, which are invented fairly quickly when the conditions are in place, are not worth patenting because the term would only be a few weeks.

Of course, there should be a cap (say 20 years?) and a public review process before the patent is issued so that if an inventor "accidentally" misses mentioning relevant prior art, a competitor can fill it in for them.
-- arvin, Nov 20 2013

Been there, done / don't do that Been_20there_2c_20done_20that_3f
[normzone, Nov 20 2013]

There is nothing new under the sun.

Fortunately the Halfbakery is kept well in the shade.
-- Alterother, Nov 20 2013


Given that most patents generate no licensing fees, and those that due typically do so in proportion to their usefulness and revenue generating capacity, I do not think that this idea sides anything
-- theircompetitor, Nov 20 2013


You've failed to grasp the patent process, [arvin].

For one thing, patents aren't (or shouldn't be) granted unless there's a significant inventive step. Something which is an "obvious thing" (as you put it) isn't patentable in the first place.

Second, if a patent describes a novel extension of prior art, and requires that prior art to be exercised in the performance of the patent, then anyone using the patent will have to pay licensing fees to the owners of the prior art. In other words, a patent that depends heavily on prior art is already less valuable, inasmuch as it may lack freedom to operate.

So, this solves a problem that doesn't exist.
-- MaxwellBuchanan, Nov 20 2013


The only reason that patents exists is to give incentive to disclose inventions, otherwise people sit on them and don't tell anyone. Every single patent law comes down to this.It's not an incentive to invent per se, people are naturally lazy and will invent because they hate doing what they are doing. I wouldn't disclose my invention in your system, cause it seems arbitrary. Take my chances with a trade secret.

For example, obviousness law is just a very necessary mechanism to prevent patent abuse, otherwise people would read patents, add two together and say "patent please." Which frankly alot of people do file for. But its not a "real thing" outside of a bureacracy managemwnt tool. Its main purpose is to create a burden to explain why some new combination of elements is more than the some of its parts. But how do you measure this in your system?

True Novelty is "solving a problem that only you know exists", ie, "oops that didnt work" you are actively and iteravely experimenting and empirically progressing by making refinements to your own problems (without telling anyone). If you keep track of these failures it's the easiest way to get a patent.

The only reason any cares about giving you a patent is because its published after eighteen months and at least the world will know about your cool idea, which is very good for the economy and other people building other things.
-- leinypoo13, Nov 20 2013



random, halfbakery