Comment by tensor | Hacker News Reader

Comment by tensor | Hacker News Reader

Parent Full thread

As a minimum test for whether something is non-obvious enough to be patented, I like to think of whether it could have been published in a computer science or UI journal. Similarly, if it has already been published in a journal, then it should be off limits for being patentable.

Let's consider multitouch. This invention has already had numerous articles written about it [1]. Thus it is patentable, but Apple is decades too late. That it can be used with any input device that supplies multiple touch inputs is obvious. You cannot write a paper about that.

Can you write an HCI paper about the disappearing scroll bars? I would be a bit surprised, although possibly. I imagine if you did, it wouldn't be a very prominent paper or in a very good journal.

In any case, I think that an invention being publishable is a minimum. There are also many publications that do not represent techniques so innovative as to deserve a short term monopoly.

Also, your distinction of ideas vs invention is not shared by books that describe good patent drafting. On the contrary, the book Invention Analysis and Claiming [2] opens with:

"Part I - Identifying the Invention - takes as its central premise that an invention is not a thing, but a concept."

[1] http://www.billbuxton.com/multitouchOverview.html [2] http://books.google.ca/books/about/Invention_Analysis_and_Cl...

> Similarly, if it has already been published in a journal, then it should be off limits for being patentable.

It is. Patents can't be issued for something available to the public; this is called "public disclosure" and has bitten lots of researchers who publish before applying for their patent.

Repeat after me: Patents do not cover ideas. Patents do not cover features. Writing about an idea or a feature in a computer science journal is not prior art.

Star Trek has the warp drive. So, if someone invents the technology to give us a warp drive, you think that it isn't novel because we've seen it on TV going back to the 1960s!

I think you mischaracterize how CS journals work. What is described there is usually the characteristics of and results from a particular implementation.

If you invent a warp drive you will be awarded a patent over the implementation of it, that was never explained in ST for obvious reasons. So if you invent a real one, built to look exactly like the Star Trek one with an "interface" very similar to the one in the serie, and try to patent it's design and "interface", then Star Trek's warp drive will be prior art (for the design and interface, not the way it works).

See, not all patents are of the exact same kind, and there's a lot of things to patent about any thing. That's the problem... nowadays things that shouldn't be patented, like designs and user interfaces (even algorithms), are.

So I'm curious, how should design be protected? Unique designs come from other places besides technology -- fashion industry files design patents on almost every "unique" accessory they produce. If we were to remove design patents, something would need to fill the void as trademark/copyright wouldn't protect physically formed objects. We could argue that it shouldn't be there to begin with, but then nobody would make any money in the high-end fashion industry.

Apples pinch-to-zoom patent covers what we call pinch-to-zoom on a touchscreen. There is no other logical way to implement pinch to zoom on a touchscreen.

Therefore, they effectively have a patent on the feature pinch-to-zoom on smartphones.

It is also obvious. Had I been working on the iPhone and decided I wanted to implement the pinch-to-zoom feature from minority report, I would have done it exactly as described in Apple's patent--there is no other logical way to do it--using a touchscreen.

I can't use a different underlying algorithm, or even a different kind of touchscreen. If I implement pinch-to-zoom on a touchscreen I am violating their patent.

I think Apple's patent is actually multitouch on a resistive touch screen. They do not have a patent on implementing gestures by measuring the speed and distance of multiple touch input points.

The later is what has plenty of prior art and the former is exceedingly trivial and really should not be patentable.

I think the problem in this case is the jury just decided to ignored prior art.

Quoting from this link: http://news.cnet.com/8301-13579_3-57500358-37/exclusive-appl...

> "He had experience. He owned patents himself...so he took us through his experience. After that it was easier. After we debated that first patent -- what was prior art --because we had a hard time believing there was no prior art." "In fact we skipped that one," Ilagan continued, "so we could go on faster. It was bogging us down."

I doubt that in the appeal Apple will be so lucky.

One legitimate line of this critique, though, is the quality of (peer?) review at the patent office. Is it doing the job at the right level?

I highly suggest you read the above quoted book by the American Bar Association on drafting patents. Emotionally repeating your own personal definition of the word idea helps no one. When we use the word idea in the context of invention, we are talking about the idea of an implementation. For example, the idea behind a steam engine, it might be at its most general: driving a piston with steam. If this idea did not have prior art, you most certainly could patent it as the idea embodies a technique.

Unfortunately, I don't have a good go-to source explaining what computer science is, but I assure you that computer science does not involve writing science fiction novels.

Reply on news.ycombinator.com