Hacker News Reader

Comment by coldtea | Hacker News Reader

>It's the "patent pending" part that sucks. They're trying to patent the idea of changing your video card's CLUT (color look-up table) to reduce eye strain, which is a fairly obvious and trivial thing to do once you know that blue light affects melatonin production.

On the other hand, since nobody else has done it (or patented it yet), real life proves that it's not that "obvious".

Like the "egg of Columbus" some thing are obvious in retrospect: https://en.wikipedia.org/wiki/Egg_of_Columbus

View on HN

I said "obvious and trivial thing to do once you know that blue light affects melatonin production." In the past people did other hacks with CLUTs (like some early 3D games baking gamma correction into the CLUT so that proper linear lighting could be displayed without an expensive postprocessing pass). If you were aware of the blue light effect (I'm not a sleep scientist, I don't know how long they've been aware of it) and had some low-level graphics programming knowledge, I don't think it would require great leaps to come up with this solution. It's just that no one else was interested in or aware of this problem.

Regardless of whether it was obvious or not, I hate the idea of extremely simple solutions being locked up behind patents, keeping the world today worse for the sake of hypothetical future innovations. It seems especially scummy the way they're trumpeting over social media about the harm that blue light causes while quietly trying to profit over an extremely simple solution to it. More than likely they know that Apple will never allow this on their app store because of the potential for abuse the necessary APIs would provide, and their end game is hoping the angry mobs will convince Apple to implement this as an official feature (and thus, pay them royalties). And more than likely, Apple would have implemented this years ago, had f.lux been an open source student research program with no patent applications attached.

>I said "obvious and trivial thing to do once you know that blue light affects melatonin production."

That doesn't change my argument. That too was known for ages, and still no light changing apps for it like flux (with the exception of a Linux app which I think came later).

>That too was known for ages, and still no light changing apps for it like flux

How is Joseph Programmer von Notasleepscientist supposed to know that a blue-light dimming filter is something that he might want before he sees the effect in action? I'll give some credit to f.lux for popularizing the idea, but once the idea is out there, there are only two ways you can implement it: postprocessing in software, or postprocessing in the CLUT. You wouldn't need to know anything about how f.lux works to come up with one of them yourself, once you are aware of the idea.

This fact is in neuroscience and psychology text books. It has been well known for quite a while. That said, it's true that others haven't done this prior to f.lux coming out. From a business model perspective, they are making a utility that could have been bundled into the OS. It's a small feature, not a whole business, unless they can gain exclusive use of their invention.

Whether this patent protection incentive was necessary for them to try to get their product out there in the marketplace is a harder question to answer. Certainly, they have chosen a business model that relies on getting the public familiar with the technology and then licensing it to manufacturers. It does seem to me that many fewer people would know about flux if they were charging for it initially, but it also doesn't seem like a huge leap forward tech wise, though it may provide a lot of value. I'm kind of on the fence.

EDIT: I did a little digging. While I happened to have known about this fact for a long time, and am pretty sure it was in my textbooks as of around 2005 or 2006, it seems this may have been discovered around 2001. https://www.ncbi.nlm.nih.gov/pubmed/11487664

Actually, the stuff about blue light and melatonin in humans was discovered by biologists in the 90s. They get all the credit here. Obvious applications of that knowledge would not normally be patentable, but the US patent office is pretty much batshit insane at this point.

In fact, a long felt and unmet need in the field would serve as a secondary indicia of non-obviousness.

Obvious or not is a bit of a red herring. The key test should be whether somebody would have done the same thing without a patent. If they would have, then a patent is unnecessary - the government shouldn't be in the business of granting arbitrary monopolies, only those monopolies that specifically promote something, for example, research. If a patent is granted for a thing that would have been done anyway, then that patent is a kind of a dead loss for society.

In the case of flux I think it's relatively clear that it would have been done without a patent, but maybe some people would disagree.

There are only two real incentives that we, as a society, have to grant patent protection. One is to ensure that inventors have a fair chance to recoup their R&D investment. The other is to offer an alternative to trade secrets, where innovative insights are hidden from the rest of us, potentially forever, by a single proprietor.

Obviously, neither of these applies here. There is absolutely no reason for us to grant patent protection to f.lux. This is so "patently" obvious that the burden of proof falls on those who would argue otherwise.

>There are only two real incentives that we, as a society, have to grant patent protection. One is to ensure that inventors have a fair chance to recoup their R&D investment. The other is to offer an alternative to trade secrets, where the innovative methods are locked up, potentially forever, by a single proprietor. Obviously, neither of these applies here.

I actually disagree with both -- let R&D happen by public research (e.g. universities) that competes for funds based on results, and then makes said results available for everybody (at least in the same country who did the paying).

But that said, I don't see why the people who did this "obviously" don't need a "fair chance to recoup their R&D investment". Does it say anywhere that the R&D investment must be huge? Because that's not the case with tons of patents -- some are just accidental inventions, like the fabled 3M's post-it notes.

>There is absolutely no reason for us to grant patent protection to f.lux. The burden of proof falls on those who would argue otherwise.

Actually if the patent office DOES grant them a patent, then the burden on proof falls on you.

But that said, I don't see why the people who did this "obviously" don't need a "fair chance to recoup their R&D investment". Does it say anywhere that the R&D investment must be huge?

Rather than haggling over what constitutes a "huge" investment, let's generalize the question a bit. True or false: we'd be better off if everything that could be patented under the present USPTO rules was patented.

If your answer is "true," we're done here.

If your answer is "false," then you agree with me that the current criteria for patent grants are inappropriate and counterproductive.

Because that's not the case with tons of patents -- some are just accidental inventions, like the fabled 3M's post-it notes.

Correct, and we need to ask what we got in return for allowing 3M to patent Post-It notes. If something is that trivial and that inexpensive to develop, then there's no need to incentivize it. We already have an institution called a "market" that's well-suited to sort out the winners and losers.

Somebody else has already done it. https://github.com/jonls/redshift has been by go-to for a few years now. It's my suspicion that Jon has no intention of patenting the idea.

This is also true (also called hindsight bias).

Reply on news.ycombinator.com