[ohyes]

> It is novel in that they couldn't interact that way before, and now they can, but that novel behavior isn't an invention its merely an implication enabled by the real work of doing the connecting. Its like patenting the idea of driving your car to Grandma's.

I'd say the proper term for it is 'discovery'.

Connecting computers together in a network and creating the infrastructure (ethernet, routers, DNS servers, stuff like that) is all actual invention. If someone creates an ethernet that is 200x better than what we currently have, or a much better way of routing packets etc, they should get credit for that invention and be able to patent it.

Once you have the physical infrastructure, anything else that you do with it is just discovering a neat property of a system which has already been created.

It is like patenting gravity, the Pythagorean theorem, or a certain chord progression. (Imagine if someone had patented IV-V-I, we wouldn't have had nearly 60 years worth of music!)

In this way, I would say that Berners-Lee didn't invent the web, he discovered an extremely interesting and useful property of combining computer networks and hypertext.

[eridius]

Inventing hard things is worth patenting. Discovering emergent phenomena is not.

[jaekwon]

If it's so hard, do you really need a monopoly enforced by statute? The harder it is, the more natural temporary monopoly you get until others start copying you.

I say, nothing is worth infringing upon a creator's right to create. Please correct me if I'm wrong.

Even if there are some cases where a patent might have been beneficial overall, you still have to weigh the benefits against the costs of having a patent system at all. How much GDP is spent on patents in general?

[eridius]

Inventing hard things is hard, but copying other people's inventions is significantly easier. Don't forget, part of the point of the patent system is to encourage inventors to publicly document how their invention works so other people can use it (with licensing), instead of keeping it secret and possibly having the knowledge lost over time.

It seems to me that if someone can describe superficially what an invention does and you can build it on the spot, then it probably doesn't deserve a patent. For example, if you were to describe the appearance of an engine block to me, I wouldn't be able to replicate the combustion engine, and it's probably worth patenting. But if you tell me that the user clicks on one button to purchase an item without going through a checkout phase, I can easily replicate that without any knowledge of how it works. Although come to think of it, that basically negates all business method patents.

[hexagonc]

Inventing hard things is hard, but copying other people's inventions is significantly easier

Copying is only easier if you can reverse-engineer what they did. Sometimes, just seeing the invention brought to fruition provides enough clues for any skilled tradesman to reproduce the parts that are unrevealed. Well if that's true, then the patent should not have been granted in the first place, like you said! Remember, ideas by themselves aren't supposed to be patentable, only implementations of ideas. You see, the patent doesn't become violated just by writing or communicating about it. The violation occurs when you build something that really starts doing what the patents describes. The problem with software patents are precisely how fuzzy this distinction can become.

Software is special precisely because it is so general and flexible. Take the 'one-click' patent that you cited above. Suppose web browsers were built with macros so that you could script them to automatically verify shipping address and payment info upon clicking "buy" and confirm the order. To the end-user, this would be functionally equivalent to the "one-click". Should this macro behavior be blocked? What if an e-commerce site distributes a script for implementing one-click functionality that can be installed by customers? Would this violate the patent? I'm not sure.

[eridius]

Copying is only easier if you can reverse-engineer what they did.

Sure, you can't always copy something just by inspecting it. For example, if I invent a way to laminate baseball cards extremely well, you're not going to be able to inspect the laminated card and figure out how I did it, all you're going to see is that it's a really great lamination. And this is where the other aspect of patenting comes into play. If I never tell anyone how I did it, and I die, then my invention is lost. But if I patent it, then I've publicly documented for the world what my invention is, and anyone that wants to use it can license it from me (and, eventually, it will pass into public domain). So basically, the patent system works quite well for physical products or processes.

But, as you say, when you're talking about software patents (or really, business method patents in general), everything becomes really fuzzy. What's the difference between a business method and an idea? Why is one patentable and the other isn't? The more I think about this sort of thing, the more I'm convinced that all software patents should be invalid, period. If I write an instruction manual, I can't patent that. Copyright is what covers me there. So if I write instructions for the computer (i.e. software), why is that suddenly patentable? The fact that all software patents need to describe a computer executing the operations (which is what turns it from an idea into a patentable process) just shows how absurd the whole idea is.

[mironathetin]

Just look at the results of patent. Large companies with money copy, despite of patents. Thats why all smartphones look the same and do the same. If its about Samsung against Apple, there will be endless lawsuits. If its Samsung against you, Samsung will go on using your patent and there is not much you can do against it. This is an old argument: patents protect large companies. It is still true.

Just imagine where the internet would be now, if every website developer had to face lawsuits from various ingenuous inventors, who happened to use his two lines of code before him. It would only be manageable to develop internet code by Google and Facebook, with enough lawyers in their backoffice.

[erichocean]

> Although come to think of it, that basically negates all business method patents.

You said that like that's a bad thing...

[bocajuniors]

The only point of the patent system "is to encourage inventors to publicly document how their invention works so other people can use it (with licensing), instead of keeping it secret and possibly having the knowledge lost over time."

[jaekwon]

Is this some documented "official" purpose of a patent?

Either way, I have a hard time believing that this is a bigger problem than managing a patent system per se.

Also, there are probably much better ways to protect secrets systematically for the benefit of both inventors and the public.

More plausible is the potential benefit of being able to broadcast an invention and attract investors without having to give up a first-to-move advantage. Elisha Gray might suggest otherwise.