Michael Risch (noted IP law professor) Visits Groklaw | Hacker News Reader

Michael Risch (noted IP law professor) Visits Groklaw

groklaw.net

23 pointsfpgeek13 comments

I think the Swype example is a great example of a bad patent. In this case essentially the patent is protecting the first person to come up with an idea (and claim it) that may turn out to be popular/profitable later on. I dont think patents should protect "whoever called it first" like calling "shotgun" to claim the front passenger seat in a car [possibly a regional thing...] This basically incentivizes looking at current tech, brainstorming "cool ideas" and a basic implementation, then patenting them so that in a few years if the idea happens to come into vogue the lawyers can be let loose.

The Swype concept is what any mild-mannered engineer would come up with when faced with the problem of "come up with more efficient ways to input text on a handheld touchscreen device". Here's an idea: "Use the QWERTY keyboard except don't lift your finger when typing hurr". I mean, this obviousness can be tested empirically. Take a group of mild-mannered engineers (ignorant of Swype) and ask them to brainstorm ways of inputing text into a handheld touchscreen. How often is the Swype concept "discovered"? I would venture every time, but hey hindsight bias and all. Furthermore, there is nothing difficult (in the sense of requiring years of research and investment to solve) or technology-expanding about its implementation in software.

I'm not sure if "requires years of expensive R&D to produce" should be the only criterion for patentability. But clearly Swype, Amazon's one-click purchase, slide-to-unlock, in-app purchasing, etc. are all poor patents and hurt the advancement of the industry and consumers.

I don't think Swype's patent is quite as general as we seem to be assuming (see below). If that's right, I think that could easily change the analysis of obviousness (though it also might not, of course).

When undocked, ASUS Transformers (and the Transformer Prime) have the option of swiping-based text entry. It is the same concept, but implementation is pretty clearly not Swype's (I've used both and, among other things, it isn't as good). Nevertheless, AFAIK, ASUS hasn't been sued over this, despite Swype's owner (Nuance) being known for a certain amount of litigiousness (see: Vlingo).

I suppose it is possible that ASUS licensed Swype's patent, but not their implementation, but that seems strange (how would the cost-benefit analysis on that work?). I think the more likely answer is that they're avoiding Swype's patent (or patents) somehow.

The really bad thing about patents like Swype and "swipe to unlock" is that they are just uses of a touch screen. Once you've got a touchscreen (which is the hard part, full of actual inventions), thinking of all of the neat ways you can use it is frikkin' easy. It may be novel in the sense that before there were touch screens, it was impossible and now its not, but I'd say thinking of an application of a technology does not rise to the level of an invention.

I always say, "Its like patenting 'driving your car to grandmas'. The car is the invention, not deciding its a good way to go see grandma."

If Swype was obvious in 2003, why weren't there multiple (or any?) slide-to-type QWERTY keyboards written for PalmOS or early Windows Mobile?

I've owned PDAs since the Palm 5000 in early 1996, and I tried all the alternative keyboards, from Fitaly and Atomik to MessagEase and.. something else like MessagEase. I don't remember anyone even suggesting the idea of a sliding QWERTY keyboard; there were alternate-layout tapping keyboards, and there were star-layout sliding keyboards, and those were the two sides of the debate. No?

I disagree. I think swipe sounds like a hideous kludge.

Look at how people type - there are fully trained touch typists, people using most of the fingers in the right place most of the time, people using two fingers on each hand, ending up with the hunt and peck.

I'd be amazed if Swipe can cope with all of those, and if it can they deserve a patent.

I think the question of obviousness is the most important question in the patent debates and the hardest to answer. By asking the question, as you've proposed, you are necessarily giving information about the solution. Knowing which questions to ask is critical to inventing. Swype figured out the right question. It wouldn't be fair to prime the pump by asking it to others.

This is not to say there aren't problems with software patents. There are vanishingly few patents that really tell you how to implement the "invention" and most of them don't actually do anything useful.

Isn't that exactly reader5000's point though? Swype might have answered the right question, but so many patents were filed at the same time, someone was bound to. It might have been a novel idea at that moment, but how much reward does the government need to bestow on something that you could essentially come up with in an afternoon?

If everyone had copied the interface immediately, I would have felt for them making no money on "their" invention, but what progress in the arts and sciences is the patent really protecting? Imagine a world where for 1984-2004 Apple had patented everything down to the "OK" button in a system message. Should someone have a right to retard progress like that? Should someone have to reinvent something slightly different every time they want to enter the market?

I think the arguments he makes are sound. I think there is value in having patents (remember, before patents was trade secrets and guilds hoarding knowledge), but there are serious costs to having bad patents.

I also agree with him that, in today's world, trying to distinguish between software and non-software patents is impossible and pointless. We will either have software parents or no patents, and the latter isn't likely, so let's figure out how to make the former work.

He is not visiting, he has been a member all along.

Fair enough. I was having a hard time summarizing what was going on. "Debating" sounded too adversarial, "addresses" didn't capture the spirit of the interaction, etc.

What would you suggest?

I would simply keep the title as is: "Curing the Problem of Software Patents". Editorializing of titles is not encouraged on HN.

I thought about that, but I felt it wouldn't accurately convey what was interesting about the link (which was not just the subject of the discussion but who was doing the discussing).

Again, will PJ please explain why her anonymity is necessary today? It's been too long with this charade, we need to know the benefactors.

Reply on news.ycombinator.com