Comment by amluto | Hacker News Reader

Comment by amluto | Hacker News Reader

Parent Full thread

Abolish it entirely.

I'm not really kidding. I think that evidence that patents are beneficial in any field is lacking.

Alternatively, a few major reforms would help. For example:

1. Decrease patent terms to just a few years in most fields.

2. Make it clear that math, in any form, is not patentable. Codec patents, for example, should never have been valid.

3. Eliminate submarine patents. Specifically, require patent holders to notify infringers of their infringement in a timely manner. If you know or should know that my specification, technology, or product is covered by your patent, you must tell me what patent covers it and why. If you give me a list of 700 patents that "might" cover it, you get some appropriate penalty. If you fail to notify me, your patent no longer applies. If my infringing technology is public and you fail to publicly disclose your patent, you cannot enforce your patent against anyone who infringes it using my technology or a derivative thereof. By "should know", I mean that, if an attentive participant in whatever market I'm in would have known of whatever details of my technology infringe on your patent, then you "should know". Even if my technology is still under development by a standards body.

#3 is a big deal. It means that owning and maintaining a patent is a bit expensive. It means that, if you patent some incomprehensible detail that might apply to future technologies, then you must tell people that their future technologies would infringe. This, by itself, would likely blow away most codec patents. And AOM would be spared the expense of their patent search. Simply publish AV1, wait the prescribed time, and AV1 is in the clear.

4. Covenants not to sue and blanket licenses do not waive requirements under #3. If an AOM member wants to give a blanket license of their whole portfolio under defensive terms to users of AV1, they still need to disclose how AV1 infringes their patents or they can't use those patents defensively in the future against AV1.

I agree with some of this, and I think most software patents are garbage and should not be legally permitted. But I'm stuck on codecs. Codecs are real, as real as a bag of apples. And they're complicated, extremely specific (and necessarily formally specified), and the result of a lot of work, sometimes brilliant work.

If there's any software that has a chance of convincing me of patent worthiness, it's probably codecs. Or at least the co(d) part. I'm muddled on decoding because it seems like just figuring out how to read something should be treated differently.

Is the term 20 years now? That's too long for any kind of software, even codecs. I think an important principle that should influence the design of patent law is the rough likelihood of the invention being made by others, or qualitatively matched or exceeded by others, in various spans of time. (Of course you have to model this assuming certain patent laws in place, as well as no patent system at all, to really do it right.)

I despise patent predators, née trolls. But MPEG-LA isn't a good example for me to work with because I think they're actual companies that do stuff – make movies, Blu-ray players, TVs, etc. Some of them invented DVD and Blu-ray probably, and evidently they invented MPEG-2, which I remember was viewed as pretty slick and technically awesome when it came out (as were DVDs – I remember VCRs and boy did they suck).

But 20 years? Yeah, no, this stuff should be 10 at most. I mean, would it hurt us to have 10 year terms on codecs? Would we get lesser quality codecs? I doubt it. HEVC is pretty amazing though, and it's clearly better than VP9, so it looks like the proprietary stuff beats open source in this field (tiny sample, yes, but there are many more examples available). Daala seemed to be going nowhere. Mozilla didn't seem to be able to finish it. Now some companies have teamed up for AV1, FOSS. I guess we'll see how good it is compared to HEVC.

> HEVC is pretty amazing though, and it's clearly better than VP9, so it looks like the proprietary stuff beats open source in this field

This is just a restatement of the fact that codec patents exist.

When the state of the art advances because of improvements A to Z and A to W are in the public domain, the proprietary codecs are still "better" because they use A to W plus X, Y and Z which the open source codecs can't use because of the patents.

The real question is whether the patented things would have been invented without the patents. But the answer is probably yes, because the industry is full of huge companies like Google and Netflix that directly benefit from improving compression regardless of whether they can sell it to anyone else.

Very good points, AnthonyMouse.

AV1 is better than HEVC, is not patented, and was invented by companies interested in conserving bandwidth out of need. So yes, things get invented without patents.

>AV1 is better than HEVC, is not patented,

Being royalty free does not mean it is not patented.

IIRC they did patent it to use as a threat against trolls, but promise never to enforce it against fair users.

AV1 is better than HEVC? I've never heard this claim. Does AV1 even exist in final form yet?

Is it going to use more battery than HEVC? VP9 uses much more battery than H.264. Even when both have hardware decoders, VP9 uses a bit more juice. It's the computational complexity difference. Of course HEVC probably uses more juice than h.264 as well, but VP9 is notoriously inefficient and it would suck if AV1 was less energy efficient than HEVC.

It is in the interests of media companies to produce better codecs to drive the efficiency and effectiveness of their primary revenue stream. I mean, look at Netflix - they have a huge driven reason to invest in codec development.

The fact that all these businesses have vested interest and need for the best possible codecs, they have plenty of reason to invest in making new ones for self-serving purposes. No patent revenue potential required.

Look at AOM - an organization formed entirely to circumvent the patent system to produce a royalty free good codec everyone can actually use. When companies look at the patent encumbered codec of the day and reject it because of the patents you have a demonstrable show of how detrimental patents are to innovation.

OK lets say we abolish patents tomorrow, and companies do what they did before - start using trade sicrets and copyrigt law. Netflix/Google/Apple/Amazon team up to create a great new codec that can only be used as a binary blob, because its compiled obfuscated etc. - something like what the game industry uses right now to protect games.

People start poking around to figure out how it works, because hackers are curious folks like that, some code gets leaked at some point, and someone starts doing codecs as open source, for fun and fame. Said companies use that os and improve on it, crating the next generation.

Now iterate this a few dosen times and what I think companies would figure out its easier to just team up and outsource this kind of stuff. Or maybe we’ll see what happened in the game engine world - with only copyright protecting the code - devs would not be able to use “the exact same code” but could incorporate new techniques at will, limited only by their brilliance.

And then there will be paid and OS ones, but the paid ones will have reasonable terms because of strong competition from other paid or OS codecs. In any case I think it would be an improvement to the current situation.

> Netflix/Google/Apple/Amazon team up to create a great new codec that can only be used as a binary blob

But the practical reality is that these companies are doing the exact opposite with AV1:

http://aomedia.org/

https://en.wikipedia.org/wiki/AV1

The Alliance for Open Media was formed because royalty-free audio and video formats save everyone time and money, particularly in the context of the complicated licensing terms for HEVC.

Netflix, Google, Apple, and Amazon all do business on the web and the W3C's royalty-free patent policy benefits everyone:

https://www.w3.org/Consortium/Patent-Policy-20170801/

It's worth mentioning that when MPEG-LA was born, the way multimedia was moved around was quite different to how it is today. Back then it was generally transferred via physical mediums so the cost of both the codec royalties as well as the storage medium could easily be passed onto the customers ("customers" in this context doesn't necessarily mean end users). Whereas these days it's more commonly provided as a service with a fee and where the supplier has to pay for the bandwidth used. This means they have an added incentive now to improve codecs in a way that wasn't there 20 years ago because by improving codecs they can now reduce their costs while charging the same amount to their customers.

> Now some companies have teamed up for AV1, FOSS. I guess we'll see how good it is compared to HEVC.

Here are some comparisons between AV1, VP9, HEVC, and H.264:

http://www.streamingmedia.com/Articles/News/Online-Video-New...

Whether or not VP9 beats HEVC, its advantages are that it beats H.264, can be used without the licensing headaches of HEVC (or H.264), and has double the decoder base of HEVC:

https://ngcodec.com/news/2017/10/21/why-we-are-supporting-vp...

You can use VP9 and Opus for video today with a view to using AV1 and Opus in the future.

That second link is remarkably hard to read. I hate this hipster web design trend of using faint text that barely contrasts with the background. That's so the opposite of what a reading experience should be.

Anyway, on VP9... its disadvantage is that it's very computationally taxing compared to H.264 (and presumably VP8). MS Edge won't use VP9 on laptops that don't have hardware decoding for it, because decoding VP9 with software kills battery life. Are other browsers following similar policies on laptops? (this is only applies when the laptop is running on battery, and Edge will let you force VP9 even on battery in its flags).

H.264 is quite elegant and efficient from a computational perspective. Note that even when laptops have and use hardware decoding for VP9, they still use more battery than hardware decoding H.264 (which will be in any computer that has hardware for VP9). Though there must be a threshold where a difference in bitrate would equalize them.

I guess the days of H.264 computational efficiency are over if we want better compression. I'll read your links to see what to expect from AV1.

> I despise patent predators, née trolls. But MPEG-LA isn't a good example for me to work with because I think they're actual companies that do stuff – make movies, Blu-ray players, TVs, etc. Some of them invented DVD and Blu-ray probably, and evidently they invented MPEG-2, which I remember was viewed as pretty slick and technically awesome when it came out (as were DVDs – I remember VCRs and boy did they suck).

To a certain extent, I agree. If license fees for MPEG codecs have kept codec researchers in jobs, then that's a good thing.

But take a look at the MPEG-LA HEVC patent list [1], keeping in mind that includes but a small portion of the HEVC-related patents. From what little I know about codec development, I expect that HEVC's advantage over AVC is from a few large changes and many small changes. However, it's difficult for me to believe that there are literally thousands of substantial patentable improvements over the state of the art in HEVC. How many of those companies are paying for cutting-edge research and how many just got a patent that mentioned coding so they could get a part of the royalties?

I have neither the time nor the expertise to go through that list and tell you which are "worthy" and which are "trollish", but its farcical to suggest that all of them are so valuable that the owners deserve to rent-seek the entire video industry.

1: http://www.mpegla.com/main/programs/HEVC/Pages/PatentList.as...

> evidence that patents are beneficial in any field is lacking

Pharma?

It's trendy to pretend that pharma patents are the Devil incarnate because they enable exploitative pricepoints by the patent-holder, but the fact is that's what enables the profit motive for the research in the first place. (At least, I've never seen a serious argument against this, only righteous blustering.)

(The fact that US pharma companies spend more on advertising than on research, has no bearing here.)

Much of the research is done in non-commercial settings. Then companies pick the most promising stuff, patent it and run clinical trials. The real cost is the clinical trials, not the basic research. Read the wording, they often talk about the "cost of bringing a drug to market" which isn't so much the discovery phase as the "prove it" phase.

As an example of exploitation see the statin drugs used to lower cholesterol. These were originally isolated from an ancient Chinese herbal "red yeast rice" and found to be effective. So some company patented the molecule, did clinical trials, and profit! I chose this one because I have personal experience getting dramatic reduction in my cholesterol from taking red-yeast-rice. It's also a nice example of how the system is broken. We all could have benefited from trials proving that red-yeast-rice reduces cholesterol, but that's not what happened. The company would not have funded the trials if they couldn't patent the molecule.

IMHO if clinical trials were cheaper there would be no need for granting monopolies (via patents) to cover the costs. I also don't see why trials are so expensive. There are plenty of doctors treating plenty of patients that are willing to try things. All we really need is a set of rules to protect people from irresponsible trials, and formal study procedures to document results and determine efficacy. It doesn't seem too complicated right?

No IP protection would destroy the pharma industry. Like you said, the cost is not in basic research, it’s in making sure that new compounds do what they are claimed to do and don’t kill people along the way. But why would anyone go through that process if a generic maker could piggy-back off of that work once it’s completed? There would have to be a redesign of the entire institution as we know it. Maybe that’s a good idea, but just eliminating patents alone would be disasterous without a lot of other changes

Researcher: "I've found that this rhinovirus is devastating to certain kinds of tumors when injected into them... in mice. We need to do more testing, and in humans."

Company: "But that's a naturally occurring virus, we can't patent that so we're not interested. Do something else, or find a way to 'make it better' so we we can patent it and make money."

True story. Now imagine this:

Researcher: "I've found that this rhinovirus is devastating to certain kinds of tumors when injected into them... in mice. We need to do more testing, and in humans."

Oncologist at major center: "We've seen a number of terminal patients with that type of cancer, current treatments don't work especially at stage 4. We could test that virus on a number of people if they were willing to try it."

Cancer Patient "Sign me up!"

There has to be a way to make this work, and it should cost far less than the current approach. I'd argue that it would cost the government less to fund the studies than to pay medicaid costs for the patented treatments. Sure, change would need to happen but it would happen fast, and in the mean time nobody would eliminate current treatments - they're already established and profitable.

An observation without any kind of control is not really good evidence. Of course that means it's not evidence of the contrary either.

However I do agree that profit is a very powerful motivator, and I see no easy way to make sure the money is better spent otherwise (although there's no reason for all research to be profit driven, publicly funded research can definitely be good thing even if only as competition).

However I get the impression that we're at the point where there are so many things that can be cured if money is no objection that money is starting to become an objection, and nobody seems to be prepared to do what is essentially putting a price on human life.

> Pharma?

There was a paper from one of the Fed branch offices about this. It turns out that the US has two independent mechanisms to help pharma companies make money: patents and FDA-granted exclusivity. It's unclear that patents are needed.

> patents and FDA-granted exclusivity. It's unclear that patents are needed.

I admit I don't know much about the latter, but what about international markets?

> Abolish it entirely.

The original argument for patents was to avoid secret sauces that die with their companies.

The reality is that companies that would use secret sauces still do, and those whose technology would be trivially reverse-engineerable patent it.

And a plethora of secret-sauce companies give up and publish to attract talent: see Apple’s move to publish about ML.

>Make it clear that math, in any form, is not patentable.

Ok but CRISPR isn't math. And the distinction between designing an algorithm and designing a machine is kind of arbitrary. If you are going to get rid of one you should get rid of the other and abolish patents entirely.

CRISPR was discovered, not invented. At least the core concepts should not be patentable under any system. And then any inventions around it done at publicly funded institutions should probably not be patentable under any system.

Companies don't really invent very much, they usually find things or take existing things and try to monopolize them with patents.

> And the distinction between designing an algorithm and designing a machine is kind of arbitrary.

The distinction isn't arbitrary. If you can do something entirely in math/software then you can also hard code it into a physical machine, but the reverse obviously isn't true. There is no calculation you can perform on an abstract Turing machine that will cause it to generate electricity or thrust or refine steel or physically prevent someone from accidentally sticking their hand in a saw blade.

I don't see the difference. The process of designing a thing is the same whether it's an algorithm or a pile of gears. Sure real world inventions can involve new scientific discoveries. But most don't, and scientific discoveries shouldn't be patentable anyway.

That won't work.

Already today many companies don't even patent stuff anymore, but claim their technologies to be trade secrets, and still manage to prosecute anyone trying to share them.

From Waymks Lidar to SpaceX tech to Google's TPUs, both is even patented anymore, it's not even published in any way.

Reply on news.ycombinator.com