Hacker News Reader

Comment by legulere | Hacker News Reader

I think I never saw a better description of what's actually wrong with the patent system today than what's offered in this article:

> The Supreme Court has said that patents should only be granted for “those inventions which would not be disclosed or devised but for the inducement of a patent.” In other words, if the inventor isn’t trading information of real value for the patent, then it’s not a good patent. Unfortunately, many software patents do just the opposite. They offer no real information about how to implement a feature, only a vaguely worded description of the feature itself. In court, vagueness is often rewarded over specificity, as plaintiffs stretch the boundaries of their inventions to cover the defendants’ products.

View on HN

The title is all you need to illustrate what's wrong with the patent system.

There's another variable that the headline doesn't capture: the legal system. The fear that you might infringe and be be taken to court is often enough to deter people from innovating. The legal fees and lost revenue can be enough to ruin a business.

Their objection of vagueness isn't good in this case I feel. A patent is directed at skilled practitioners of the relevant arts, this document has sufficient detail to allow a programmer to implement the idea. Claims must define an area of technology otherwise patents would just be worked around and would provide no useful monopoly.

However it does appear to be anticipated by prior art mentioned elsewhere on the comments here. Given that, I'm surprised the patent wasn't challenged: that suggests it either wasn't useful to others or that the USA court system prevents lawful challenges to anticipated patents somehow (eg by burdening appellants - appealing against the grant - with costs).

> A patent is directed at skilled practitioners of the relevant arts, this document has sufficient detail to allow a programmer to implement the idea.

I haven't read the patent yet, but how specific is it? Because while it is one thing for the description to convey the general idea so that a skilled practitioner can make something similar, there's a big fat question of ambiguity - did I actually make the thing this patent describes or not? What exactly is the thing that's covered?

Otoh, being too specific allows someone to change one thing and claim it doesn't infringe.

That's how it should be.

Well, imagine apple puts all information necessary to build an iPhone into a patent. Then Samsung builds a phone identical to the iPhone, except it's green. If the patent specifically mentioned black, gold, but not green, then a green phone might not infringe.

My point is that too specific can also be a drawback.

Too specific is only a drawback to the patent holder. And a single-patent iPhone patent would cost a fortune to file anyway. Having overly specific patents is not remotely a drawback to the rest of us.

If apple is discouraged from inventing an iPhone because they won't be able to defend the patent against minor changes, that's a loss to society.

If you concede that patents in general are good, then you should also agree that sometimes, it's better for a patent to be general.

I'm more in the "everything is a remix" camp, at least as far as software and design patents are concerned. If Apple didn't create the iPhone, someone else would have made something similar.

Patents have two main sections that are very different from each other: 1. The claims set out exactly what inventions the patent protects. They are the part that must be novel and non-obvious. When writing a patent, you want your claims to be as broad as possible (while still satisfying the novelty and non-obviousness requirements) to maximize the protection that the patent offers. 2. The specification teaches a person skilled in the relevant art how to make/use the claimed inventions. This section includes specific details about possible implementations of the inventions. This section can discuss many different variations, and should make it clear that these are only some of the possible variations, not an exhaustive list.

Using a simplistic example based on your child comment, a sentence in the specification might state "The phone may be black, or gold, or any other suitable color." (In reality, the specific color is not likely to be addressed at all unless it plays a role in the claimed invention.)

So you're saying there's absolutely no harm in being specific, because that's a different section? That's new to me, and if correct, I retract my claim above. Thanks.

Only because the system already allows for patents that shouldn't exist because they aren't novel enough.

See, for example:

http://www.let.rug.nl/usa/presidents/thomas-jefferson/letter...

Jefferson was clearly dead set against the situation you are describing -- granting patents with only incidental modifications to the system. But then again he would be rolling in his grave if he saw (and had the modern knowledge to understand) the completely inane, obvious (to practitioners of the art) shit we've been granting patents on since about the late 1980s.

Reply on news.ycombinator.com