Hacker News new | past | comments | ask | show | jobs | submit login
The licensing program for MP3-related patents has been terminated (fraunhofer.de)
469 points by petercooper on May 1, 2017 | hide | past | web | favorite | 269 comments

Recognizing this adds legitimacy to software patents. Don't. There have been free and open source MP3 decoders for many years now.

Not believing in software patents is unlikely to change a court's mind. Instead, support formats that are royalty free from the start - Vorbis, Opus, and Codec2, for example.

The major trouble with patents, and the thing that puts them in a special category of nuisance all of their own, is that it's fundamentally impossible to know that a thing is unencumbered until each particular claim to the contrary has been hashed out in court, at ruinous expense. And there is no legal proscription against patent owners letting you think you were safe for years, and then popping up to cream off the fruits of your work. We know MP3 was encumbered. We don't, and can't, know that MP3, Vorbis, and so forth aren't encumbered now. All we can know is that nobody has popped up to hold the industry to ransom... yet.

I'm really glad New Zealand decided to disallow software patents. I wish more countries would follow suit.

I'm really glad New Zealand decided to disallow software patents

No, that is not the case.


What the Zeds have done is really not so different from what the Supreme Court decided in Enfish:

  claims purporting to improve the functioning of the
  computer itself, or improving an existing technological
  process might not succumb to the abstract idea exception
Abstract software is excepted (disallowed) but software which really controls machines is allowed.

This source is a bit clearer:


"Effectively, if the invention lies purely in software, then it is considered unpatentable. However, as outlined in the explanatory note of the SOP, patents will still be able to be granted for inventions that

a) make use of computer programs where the contribution lies outside of the computer, or

b) affect the computer but is not dependent on the type of data being processed or the particular application being used.

Essentially embedded software is patentable where it improves the operation of hardware, along with software that improves the way a computer runs. Again, this is consistent with the UK."

Also Zeds is a really strange name for us. Correct term is Kiwis or Rugby Overlords :)

Has any action been taken against them by other countries or even private industries for this?

IIRC one of the things that the TPP was supposed to do was "harmonize" (read: ram the US idea of down the throats of everyone else) software patents and other relevant aspects of IP law that were near and dear to major US companies.

Since TPP is dead, the IP-dependent industries may start looking for a new venue, as they appear to have been putting a lot of eggs in that particular basket. But there may be a short-term reprieve for countries not currently toeing the US party line.

"it's fundamentally impossible to know that a thing is unencumbered until each particular claim to the contrary has been hashed out in court, at ruinous expense"

That applies to MP3 (and all of MPEG and other 'licensed' systems) as well. See for example Sisvel's MP3 related patents that led to raids on MP3 players at the Cebit fair.

>it's fundamentally impossible to know that a thing is unencumbered

One noteworthy exception, which plays a much bigger role in slower-moving industries, is that an exact copy of something sold over more time than the pendency of any patent in existence (mostly less than 23-24 years) is clear. But the larger point holds. Comparing patent claims to products and prior art is hard and expensive.

It's not even enough that the thing you're selling be an exact copy of something old. Your manufacturing process or other business processes (sales channels, support channels) could also be infringing on some newer patent.

Can't we take winamp's initial release from April 21, 1997, add 1 year for late filing, and say anything it did is unencumbered by April 21, 2017? (Let's ignore patents filed before June 8, 1995 and issued later, since we're approaching the end of those, I hope)

If you can demonstrate standing and pay a lawyer a ton of money, then yes, you could argue that.

WinPlay3 was released by Fraunhofer themselves in September 1995.

If an implementation of a patent exists, and that patent expires, it is impossible for that implementation to infringe another patent for the same thing, because any other patent that applies would have been invalid when issued, because of the previous patent.

Eh, not necessarily, I've heard that one thing patent owners do is keep the thing ticking over by patenting something else sufficiently close to the original that they overlap and sufficiently different that they can be argued as distinct. Sure it's infringing - but it's infringing their own stuff. So, not invalid.

Note: as to whether any particular tactic like this stands up in court or not, you'll find out if you fight it in court. You and your huge heap of cash, that is. Don't have a huge heap of cash? Then the question is moot, isn't it, peasant?

Not necessarily. You could patent an improved engine. 19 years later, I could patent an improved alloy that someone might use in your engine. Your patent expires in year 20, but they still infringe if they make your engine with my alloy for the next 20 years.

How do you "support a format"? AFAIK nothing will convince Apple to release music in a better (patent-free) codec, much less convince others to dump DRM.

> How do you "support a format"? AFAIK nothing will convince Apple to release music in a better (patent-free) codec, much less convince others to dump DRM.

Then you start by not giving them your money. It's a uphill battle, but someone has to fight the good fight instead of just throwing their hands-up.

> someone has to fight the good fight

That's a noble, if politically naïve, attitude. Societies, like Congress, have a limited bandwidth for problem solving. The way we prioritize what gets solved and what doesn't, as well as how to solve it, is "politics". (This is more a matter of definition than quibbling.)

There are various political institutions, and individuals have varying degrees of influence over them, depending on how one's society is structured. But unvaryingly, attracting people with political capital to your cause is paramount. This is often done by showing such people an agglomeration of individuals supporting your cause, individually with less political value than them but collectively something interesting.

Even the people at the bottom of the chain have limited bandwidth. This is why most systems default to delegation, particularly at scale. Single-handedly refusing to support something you think is important without patching that into a political system is fine from a personal or moral perspective, but it shouldn't be expected to actually do anything. As such, it's less a good fight than good fussing about.

There's a wide range here. Complaining on a message board where most people agree with you is basically zero effort. Devoting your life's work to it is a ton of effort. There is a middle ground here.

My general thought is: you should, on a regular basis, feel at least a little bit of pain due to your opinions. If you don't then your opinion is probably worthless to society.

> My general thought is: you should, on a regular basis, feel at least a little bit of pain due to your opinions. If you don't then your opinion is probably worthless to society.

I've reread this sentence a few times... and it is actually quite deeper that it first seemed on a more superficial read. Thanks a lot for your comment.

Here's a reform to the copyright law that just occurred to me. If you buy a copy of a published work, then you're allowed to copy and distribute it as you see fit, provided that you pay the creator the same royalty that they were paid for the copy you bought. Other restrictions such as moral rights would still be in force.

That seems to get a lot of incentives right. Publishers have to pay artists decent royalties, and proprietary file formats can't be used to suppress competition. What would go wrong? Apart from the fact that this would have to be enacted over the cold, dead bodies of a lot of music company executives: that's a feature, not a bug.

I think you are confusing the issues. TFA is about the license fees paid to Fraunhofer for the MP3 codec, for which they owned the underlying and now-expired patents.

MP3 is not an encrypted codec, the format itself is quite open and doesn't place any real restrictions on redistributing music -- in fact, most of the original music sharing sites like Napster traded exclusively in MP3. The use of MP3 has no bearing on copyright or music piracy.

The real issue is that software developers had to pay a royalty to Fraunhofer to distribute copies of MP3 players or encoders. That made it impossible to incorporate MP3 into a free (as in speech) OS like GNU/Linux.

> The real issue

There's real a higher-order issue, too. Why did anyone want GNU to incorporate an MP3 player in the first place? There is no law of physics that requires songs to be encoded as MP3.

> Why did anyone want GNU to incorporate an MP3 player in the first place? There is no law of physics that requires songs to be encoded as MP3.

Dude, in 2006 ALL MUSIC was mp3.

> Societies, like Congress, have a limited bandwidth for problem solving

And as we know, due to Nyquist, N Congressmembers can at most be solving N/2 things at any given point in time.

Your right and it works. People used to say the same thing about Linux and open source software.

I wish that worked here. You'd have to establish an open source media group with direct hooks into the i* platform. Then you'd have to get major acts on that platform.

No one is making you use that platform?

That approach is a complete loser. 99.999% of people couldn't care less, your voice will be nothing but noise.

Do you suggest I don't listen to music, or are you prepared with an alternative to apple?

Why do you need Apple to listen to music?

There is radio (online and in the ether), there are alternative platforms (e.g., Bandcamp), concerts and other live venues, (second hand) physical media (e.g., CDs or vinyl), open access music, etcetera.

On top of all that there is the big middle finger of 'piracy' if you are so inclined. For many this is a morally acceptable alternative to a rotten system, for others it may remain an imponderable transgression of law and or ethics — either way, it is a viable alternative in addition to the above.

I'd love to support bandcamp but then they removed their API.

Amazon sells music and lets you download DRM-free mp3s.

So does Apple... the call here is to use something other than MP3.

Apple doesn't use MP3, they use AAC (.m4a), which is technically superior to MP3 but also patent encumbered.

They are not patent free, which is the point at hand. Do they allow ogg/vorbis or other unencumbered formats?

As an engineer? You can influence design decisions and standards for the products your company makes.

Somebody at apple decided AAC was better than MP3 because of it has better licensing for distributors, they could have gone further to consider codecs which had liberal licensing towards codec developers/manufacturers.

AAC is also higher quality than MP3 at medium bitrates[1], but this is largely irrelevant for high-bitrate iTunes downloads.

[1] http://listening-test.coresv.net/results.htm

You can only say that because of the astonishing improvements to MP3 encoding over the past fifteen years. Back in 2003 when Apple opened the iTunes Store, AAC was substantially better than MP3 at all bitrates, most particularly the 128 kbps rate Apple initially used.

In fact the decision would have been made back in 2001 or earlier, given that the original iPod released in 2001 supported AAC decoding.

(The music labels also forced Apple to encrypt everything, therefore there was no incentive to use an open format. Had they used encrypted Vorbis, that would have been seen as a giant middle finger to open source.)

I don't think the iPod supported AAC at its initial release. Virtually no software supported it at that time -- that didn't really come until 2003, when iTunes 3.0 added support for AAC (including encoding). The Music Store was introduced in the same release, so there would have been a required software update for iPods at the same time, to add support for Fairplay encrypted files.

Correct, not supported at the point of release, but its specs did strongly suggest it was factored into the hardware design.

iTunes originally sold files encoded at 128k, where AAC made a big difference over MP3.

They have now switched to 256k, where all lossy codecs are largely equivalent.

Aac is higher quality on all bitrates, it's a much more modern codec utilizing techniques not available in mp3. Using MP3 should be discouraged, it kills hihat clarity and low bass for example.

>How do you "support a format"? AFAIK nothing will convince Apple to release music in a better (patent-free) codec,

Simple: don't give any money to Apple for music in patented or DRMed formats, or to anyone else who tries to sell music that way. If you really want, you can write them to tell them why you're not buying their crap.

Companies get away with this behavior because consumers let them, and reward them with their money. You can make the argument that consumers don't have enough power for necessities like internet service (because of local/regional monopolies or oligopolies), but this just isn't true for luxury goods and services, which is exactly what the Apple music store is. You don't need downloaded music, there are other providers out there, and you can always buy it on CD (frequently for less) and rip it yourself into whatever format you prefer, like Vorbis, Opus, or FLAC.

That's exactly what I do: I buy the physical media and since this doesn't limit my rights in how do I store the music contained inside, I just rip it off that physical media and keep it safely on my machine. That way I also give all the corporations the finger and I am hopeful that even if several tens of thousands of other people do the same, they'll notice a lack of growth at least and reconsider.

I don't seriously believe the corps would actually change their ways to benefit the customer -- but this is my rebellion against their business model and it's an exercise of the small power I have.

Plus I don't break the law. I simply actively don't feed them money in their scheme which I feel are very unethical and greedy.

If buying from iTunes, you can choose ALAC, which is royalty free.

Patent-encumbered audio codecs are popular at this point because of network effects, not technical superiority. There are many ways to reduce the network effect, even without changing music buying habits. For example If you're a website designer, include a royalty-free audio format as one of your <audio> srcs. As a bonus, you'll save some bandwidth.

How do I "choose" ALAC? I thought only select albums were available losslessly.

Support Bandcamp. Support Databeats (storefront for many drum 'n' bass labels). Hell, support Beatport and Juno, even though they insist on huge markups for WAV/FLAC.

I'm glad Bandcamp and even CDBaby support FLAC for the same price. I feel like it's the best format for balancing quality vs size, and it's more open than ALAC.

I'm loving Bandcamp so much. Buy an album once, you get FLAC, plus any other formats you might want (yes, you can always encode FLAC to anything else, but the convenience of just downloading the oggs is nice). I can't imagine buying music from anywhere that doesn't do this now.

The same way I do (and you do): get both ALAC and FLAC downloads from Bandcamp. ALAC for now, FLAC if you want to play your music on systems that won't run iTunes.

Isn't MP3 a royalty-free audio format now that its patents have expired?

As I understand it, this is the year that the last of the MP3 patents expire, nd some already have, but there are still two left. One of them expires on August 29, and the last one expires on December 30.

It seems kind of odd to do this so late in the patents' life, though. I understand the general theory behind a last-minute cash grab, but you generally can't take "last-minute" quite this literally. Why wait so long?

Wikipedia says that MP3's last patent expired in the United States last month:

If only the known MP3 patents filed by December 1992 are considered, then MP3 decoding has been patent-free in the US since 22 September 2015 [...] If the longest-running patent mentioned in the aforementioned references is taken as a measure, then the MP3 technology became patent-free in the United States on 16 April 2017 [...].


Wikipedia also says (unless I'm missing something) that all the MP3 patents have already expired everywhere else too. If this is true, then Fraunhofer is ending their licensing program just because they don't want to bother getting people to pay for something they can now legally have for free.

So don't buy music from Apple, or from anyone else that sells it in non-free formats

Do you have a suggestion? Where can I find Rihanna in Ogg/Vorbis?

The simplest thing, IMO, is just to buy your music on CD and rip it to a free format yourself. Some indie artists release their work online in free formats, but for mainstream artists a physical copy is often the only way.

In the UK format shifting was briefly allowed but media corps objected and the Tories decided it was better for society not to be able to rip CDs you own. iTunes, f.e., is a tool for contributory infringement in the UK.

And pay like 10x more + shipping because you only wanted 1 song on the CD?

I'm not saying it is ideal, but the used CD market is hugely depressed right now. I buy a ton of used CDs at various downscale businesses and rip them into iTunes with Apple Lossless. Some of the stores will run them through the disc cleaner/polisher as a matter of course and they rip fine.

I'd love to be able to buy DRM-free 24/96 lossless versions of all the music I actually want but buying used CDs for under $2 apiece seems like a reasonable compromise to me. For the moment if I can't find particular albums on CD in the used market I either buy them from eBay sellers or do without. The remaining chain retail vendors for CDs such as Best Buy have a pretty hopelessly bad selection these days.

Buy from Bandcamp.

Support indie artists and buy music in MP3, FLAC, Apple Lossless, AAC, Ogg Vorbis, WAV or AIFF.

The question was specifically "Where can I find Rihanna in Ogg/Vorbis?" Rihanna is not on Bandcamp, so this is a pretty poor answer. You're essentially telling people to like different music than they do.

You can buy your Rihanna in FLAC or AAC on 7digital.

Great tip, thank you!

Spotify uses Ogg Vorbis.

and DRM. Therefore it doesn't help you that they use Ogg Vorbis internally.

Piracy scene, get the flac version and reencode to vorbis.

iTunes was one of the first online music stores to go DRM free, and even now I believe most of it is, but I prefer Apple Music for the much wider variety and greater convenience, especially for Japanese music.

I try to always buy hardware that supports FLAC (which is what everything new in my music collection is ripped/encoded in and what I always download from Bandcamp. Even CDBaby supports FLAC now, which is surprising, but must have been a demand for it).

When looking for a head unit in my car, I specifically made sure I had something that could play FLAC and oggs (and mp3s .. and aac .. my music collection is a mix of legacy formats).

Support software and services that supports that format, and try to avoid supporting software and services that do not support it. Including your own if you are a developer, of course.

Or be based in Europe where software patents are basically unenforceable.

Or make your patent-dependent development OSS.

Or just keep using MP3 and ignoring these idiotic patents. This has worked for me for 20 years; why would I do anything differently now?

You are paying for the patents whether you ignore them or not.

Software patents are banned in Europe, so the whole shebang is really only relevant in the US. While the patent offices in Europe allowed software concepts to be patented, these patents are not enforcable.

The situation is a lot more nuanced than that. After all, this is a German company holding mp3 patents. While software itself is not patentable in Europe (or, for that matter, in the US; they argue that it's not software but a process or a "method and a device" that they are patenting), there is a lot of wiggling around the law that results in software patents still being applicable.


they mader sure to get rid of this option in the obama administration.

now it is first to file. I can just patent all those formats and troll everyone until a court finally decides to consider prior art correctly when someone finally bites the bullet on a defense instead of deal.

This has been confusing me for a long time. As far as I know, it's clearly stated that software is not applicable to patent.

> it's clearly stated that software is not applicable to patent

Given that the United States Patent and Trademark Office has been giving software patents a good deal of thought (not to mention issuing software patents) [1], I'd say the situation is anything but "clear".

[1] https://www.uspto.gov/about-us/news-updates/examination-soft...

In the US, patents on ideas, business processes, and algorithms are not allowed. However, there's a goofy (court approved) legal idea that software transforms a general-purpose computer into a single-purpose computer, and that transformation makes software patentable.

The general-purpose computer to single-purpose computer transformation cracks me up (and makes it obvious that the law doesn't necessarily have any relationship with reality): "by loading this program, a computer able to do many different things becomes a computer capable of doing a specific thing."

>that transformation makes software patentable

This isn't true in many cases after the Supreme Court decision Alice v. CLS Bank in 2014. Now we have a rule more like that in Europe, where some software is patent-eligible if it's sufficiently technical (e.g., something like RSA, or better cache management) and not patent eligible if it's not technical (e.g, CRUD apps, or new views on a database).

Thank you for the update. I wasn't aware of the Alice decision.

When I said "court approved" I meant Federal Circuit, as in Alappat ( https://en.wikipedia.org/wiki/In_re_Alappat#Majority_opinion ): "We have held that such programming creates a new machine, because a general purpose computer in effect becomes a special purpose computer once it is programmed to perform particular functions pursuant to instructions from program software."

On the one hand, several Supreme Court decisions appear to leave less room for software patents. On the other hand, I don't think the Supreme Court has an overall goal in mind. I believe the problem comes from the fact that the law prohibits patenting natural laws, but allows patenting the application of natural laws. It's hard to distinguish between those two ideas; and I believe impossible to distinguish between them when it comes to software.

> As far as I know, it's clearly stated that software is not applicable to patent

This is the beauty of intellectual property experts: they can reverse a very root concept, by changing few commas, and over-interpreting.

TL;DR: Software patents are forbidden (especially in Europe) "as such". It means that if you attempt to patent the "linked list" in software programming, this will be refused. But you can patent a "method of linking a set of abstract data together, using a numerical identifier as gathering medium". This is not rephrasing: what you are attempting to patent is NOT software - you can have a set of paper sheets, with a number on top of them, and form a linked set of data (printer documents). You could ALSO have a linked list in a computer program. But what you are attempting to patent is a more general concept, that CAN be represented as a piece of software. Of course, what you REALLY want to patent is the later.

Yes, yes, this is fraud. Legal fraud. Move along, nothing to see.

It's not that clear even in Europe (which kinda forbids them, but sorta allows them). And it differs by jurisdiction too (e.g. in the US).

You can get a patent over software that is sufficient non-abstract. Where the line is drawn is extremely unclear. But there are valid software patents.

You can be free and open source and run afoul of software patents. You can infringe on a patent completely by accident. It's not like copyright where you have to prove an intent to copy. The worst thing about patents is that independent discovery is not a defense against infringement.

So you don't believe in software patents?

Software patents are bullshit.

A patent on an invention that many people would easily develop independently of one another should not, it is argued, be granted since this impedes development. [1]

1. https://en.wikipedia.org/wiki/Software_patent#Obviousness

There's nothing unique about software that makes it more likely to be developed independently by many inventors, the same happens often with hardware patents. And the same argument applies - the patent system as a whole impedes progress as much (or more?) than it advances it.

The argument for why software patents are different is premised by why patents exist.

They exist to grant someone a monopoly in exchange for them sharing their invention with the world. If I patent a dishwasher that does not clog, I will be rewarded for my invention... But once the patent expires, a skilled engineer will be able to look at my patent, and build a dishwasher that does not clog - in far less time then it took me to build mine.

The overwhelming majority of the time, Software patents grant someone a monopoly, in exchange for them sharing nothing remotely useful to society. Looking at that software patent will not save another engineer any time whatsoever, in cloning my product.

With that kind of patent, I am using the legal system to enrich myself, without giving anything back.

Not true for compression patents.

Not all software patents are alike. Some of them actually demonstrate an original, complicated, novel, non-obvious technique, in sufficient fidelity that an engineer, looking at the patent filing can reproduce the secret design, with far less effort then it would take them to do so without the patent. That is an example of the patent providing benefit to society - that merits society providing it with legal protection.

Most don't, though.

I would argue that all software patents should include complete source code and build details necessary to compile a working version.

Something like MP3 or h265 represent nontrivial investments in information theory, and I'm basically okay with patents granting a (temporary) exclusive licensing period in exchange for sharing the practical code and theoretical work.

Patenting something like "one click ordering" or "facilitating inter-personal communication by sending audio over the internet" is just ridiculous.

With your ideas about patents someone should do it out of the kindness of their heart I guess. Is that what has driven most technological advancements of the human race?

In your software-patentless world, why would a genius choose to work in software as opposed to another field inventing patentable tangible objects where he or she will be monetarily rewarded for it and able to make a living? Don't say you are already paid by your employer to program. Someone inventing physical objects would probably also be already paid by their employer to work. This brings us back to why would a genius choose to work in a field that won't reward them for their genius?

No, you get a patent for your secret anti-clog dishwasher, so that the courts grant you a monopoly on it. In exchange, you have to give something back - specifically, you have to reveal your secret design to society - so that it is useful to someone in the future. That's the contract that you agree to, when you patent something.

If you don't want to reveal your design, you could always keep it a trade secret.

If you're a genius working on software, you shouldn't have trouble filing a patent that would be useful to another engineer, once it expires. For some reason, though, people instead patent stupid shit like 'a software system that does _______.' They don't actually explain how to build a software system that does _____, they don't give anything back to society, and are instead leeching of the largess of our legal system.

If you want protection for your invention, tell us how it works. It's a pretty simple concept.

But does mp3 fall in the category you describe? I agree with you that many software patents should not be granted, but the distinction should not be between software/notsoftware.

Software patents aren't particularly useful for making money with software. Hackers produce code, which is protected by copyright, and that's how the law protects them from having their work stolen. Patents are a bad match for software, and are mainly used for two things:

1. Trolls use them to seek rent from independent creations.

2. Companies that are not purely trolls but need a vector to attack another company sometimes use them in the same way trolls do.

3. Entities that actually make software use them as a deterrent against patent claims.

Nobody actually depends on them for the profitability of software they've made, and nobody actually benefits from the disclosures in expired patents. Software patents are purely weapons. The hypothetical genius you're talking about would not be harmed at all if they went away, and would be better off because they'd be spared the vicissitudes of patent trolls' attention.

There are alternatives to patents - for example hosting research intensive parts of your system in the cloud.

This is pretty much what is happening with speech recognition these days among other things.

No need to worry about opensource, or the fragility of patents, when you keep your technology a secret.

Another approach is to use trade secrets - you encrypt parts of your application - reducing performance - enabling you to sue anyone who decompiles it in order to replicate, or indeed interface with it.

Patents would actually seem the least of these evils, if they were only less broken.

What about 4: research institute that invents an audio compression method? The work that goes into mp3 isn't just coding the method.

Debating software patents with silicon valley types is a losing battle. They're the types who would stand to profit from someones non-obvious software invention not being patentable, and so most such people are unsurprisingly against software patents. There's little nuance involved beyond that.

What is a "Silicon Valley type"?

People who see themselves as entrepreneurs/hustlers/hackers, in contrast to scientists/inventors/researchers.

Can you name any examples of "non-obvious" software "inventions" that require patent protection?

RSA is the usual example. Do you disagree with it?

Yes, I disagree. Why does/did RSA require a patent?

Require is the wrong word. The question is were we better off with or without them having a patent. Not keeping RSA a trade secret seems like a win for society.

That presupposes a lot. Who says it would have been kept a "trade secret"? And, supposing it would have been, who's to say someone else wouldn't have come along with an alternative open solution?

The difference is that person now has to sell their idea packaged in a finished product to see any return on their idea. If you could patent it and license it, you could make money on the idea while leaving the work of developing a polished product to the experts at delivering products to end users.

Ideas in the software world are essentially worthless. Execution (your ability to implement the idea) is what actually matters. So, forgive me if I'm not losing sleep over the person who has to actually do work to see "return on their idea."

The same can be said about inventions in any sphere. But that's exactly what patents correct for; to give value to ideas independent of them being productized. And so your point is not at all an argument against software patents.

Patents are not meant to be a way to stake a claim on an idea (https://www.legalzoom.com/articles/can-you-patent-an-idea) — they're a system where you get a temporary monopoly in exchange for sharing a novel, non-obvious and non-abstract invention.

To be clear, I didn't say that's what patents are for (i.e. their purpose), but what they correct for, i.e. what they do in practice.

But I don't see that the two descriptions are different in practice. Patents incentivize publicly releasing an invention instead of keeping it a trade secret (or letting it languish). The problem this is solving is that ideas aren't valued by the market, only products. But society benefits from ideas independent of products. And so the solution is a system to put value on ideas, ergo patents. This meshes with my alternate description.

Patents don't purport to offer protection for ideas.

Was something unclear about this sentence?

>To be clear, I didn't say that's what patents are for (i.e. their purpose), but what they correct for, i.e. what they do in practice.

The problem is software isn't tangible. Something like an algorithm is different. It's mathematically provable. The vast majority of software patents are based on preexisting concepts such as the patent IBM was awarded for "Resolution of Abbreviated Text in an Electronic Communications System", which is a database that looks up abbreviations like netspeak and fills in the full word. Or Google's patent for "Variable User Interface Based on Document Access Privileges" which shows something different on a webpage based on your geographical location. Something websites had been doing for years. These frivolous software patents by major tech companies are predatory and stifle innovation. They know the patents won't hold up in court, but they can afford protracted legal battles startups cant and use that leverage to strong-arm smaller companies.

Right, it was only after the patent office was invented that people started inventing things. Before that, nothing of consequence was created by humanity.

Isn't it odd that they were able to invent the patent office without an already existing patent office? Quite a miracle.

I think the landscape for software patents is worse. For actual physical goods I've seen many example where the patent system does what it intends but I haven't really seen a similar benefit in software. Maybe there is one and I'm just not exposed to it, though.

But the vast majority of US software patents are not patentable in Europe. If the US patent system would impede progress the effect should be a stronger European IT industry. I am doubtful of that though.

One still stay away from stuff that is patented in the US as a European company though, most European companies want to go to market in the US sooner or later.

That's actually a pretty great argument that software patents are a good idea for the US, it prevents foreign copycat competition from soaking up the market on us.

The crankshaft used to be patented. James Watt had to design around the obvious implementation when building his steam engines: https://en.wikipedia.org/wiki/Sun_and_planet_gear

Software patents historically have been awarded for incredibly vague, "could-apply-to-nearly-anything" systems. As a result, there have been a lot of abuses of software patents by companies, both big and small.

Software is already covered by copyright. Granting it parent protection as well allows software companies to double dip in a way that no other industry can.

Copyright and patents protect entirely different kinds of software. Copyright protects software that's more work to write than to think about (e.g. Windows). Patents protect software that's more work to think about than to write (e.g. PageRank or whatever's inside Google's self-driving cars).

Why not grant patents to the self-driving car as a whole? In Canada, you cannot patent software unless it's within the scope of a physical product, like a car.

Also, PageRank clearly doesn't need a patent. Google's done just fine against competitors that have ripped off their technique.

> Why not grant patents to the self-driving car as a whole?

How do you do that?

The first person to think of the concept of "self-driving car" gets the patent? That will just lead to a flood of patents on things that don't exist yet -- basically a patent troll's dream.

Perhaps the first company to build a working prototype? What happens to all the other companies that almost have a working prototype -- do they now have to sit on the sidelines for 10 years?

How do you even know what a "working prototype" is in that scenario? Is it acceptable if it is just capable of driving on straight roads during sunny days with no other cars or people around?

I think patents on specific, narrow pieces can make sense: for example a laser-based 3D vision system, or a specific algorithm that identifies people about to walk in front of the car (even that is tough, because there's a thousands of very distinct approaches to that problem).

Patenting the entire system (the "self-driving car") makes about as much sense as say, Herman Miller patenting the concept of a wheeled office chair. They can patent elements of their lift mechanism, specific wheel designs, or arm adjustments, but not the concept of combining wheels and a place to sit.

A very large number of patents also protect Windows and its component parts and methods such as the file system; when other OSes have tried to be compatible in any way, Microsoft has demanded and gotten royalties, for example from Android phones.

This is a nice concise explanation, thanks!

Well, copyright covers an implementation of software, but a clean-room rewrite can get around a copyright of software. In that way, no, copyright is not a real way to protect software.

And why should a cleanroom implementation be an infringement? It is not a true derivative copy. Either way, software is more akin to language or math, certainly not mechanics. We have copyrights for language, and narrow patents on applications of mathematical formulas (which in itself is debatable in its justification).

Copyrights are sufficient. You shouldn't be able to tell someone they can't use a certain software algorithm in their own program, in a different language, in a different domain. That stifles innovation.

That's what ought to be correct, but the recent decision against Facebook on VR cut directly against this view - as I understand it unavoidable mathematical and logical similarities were treated as if patented procedures. Similarity was the only thing considered. Which might force companies to the immense expense of "clean-rooming" almost everything they create to avoid problems. And that might not be a guarantee; patent law is very badly messed up by now, and Congress has refused to get involved in many issues, such as patent misuse (since the current confusion favors large companies/political campaign contributors who can best afford large legal teams and long trials.)

By prohibiting a clean-room rewrite you're skirting dangerously close to attempting to prohibit independent reinvention. That is, without hyperbole, one of the most abhorrent ideas I've ever seen.

Thankfully, software patents are unenforcable in most of the world and willfully ignored by large parts of it.


Your analogy doesn't quite fit. Copyright covers both the source code, which is more analogous to the schematics than the manual, and the binaries.

If a physical machine is not under patent protection you can strait up copy it and start manufacturing your own version. With software you have to go through an extra "clean room" step to get around copyright. In practice the clean room reverse engineering process is rarely utilized because it is expensive and fraught with the risk of a lawsuit from the copyright owner.

Yes, this is a major issue in copyright law. In practice, there is no distinction between the patent-eligible invention (the machine) and the copyright-eligible "creative work" (the manual, or the code, in this case).

That means that copyrights can be exerted in very menacing ways, like requiring users to bypass a copy protection mechanism in order to do required maintenance, as John Deere has done. That enables John Deere to pursue anyone who doesn't use an official dealer for maintenance under the DMCA's measures criminalizing the circumvention of copy protections.

Disclaimer: Not a lawyer, probably wrong.

Saying that there is no distinction isn't correct.

The scope of a patent is much broader than a single coded implementation. The copyright doesn't protect the functional aspects of the code, it only protects it as a written work. However, companies have managed to twist the concept of infringement via reproduction, and bypassing of copy protection.

Accessing a copy of code stored in memory is now considered creation of an unauthorized reproduction, bypassing copy protection under DMCA, and/or the like.

This type of protection shouldn't exist apart from preventing partial and wholesale copying and redistribution, paralleling the treatment of novels, etc. If I own a novel, I can do whatever I want to do with that book. I can write on it, annotate the margins, remove pages, make photocopies of parts of the book for personal use, etc.

With code, I can't do this? Why?

Yeah, I agree. That's what I'm getting at; in the physical world, there is a bright line between "an invention" and "a copy". You buy a reproduction of an invention, or a physical reproduction of a copyrighted work, and that reproduction is a discrete unit owned by the purchaser and not really subject to the limitations of IP (only insofar as the fundamental components are directly copied to deprive the inventor/author of the proceeds).

In cyberspace, many things that should be considered "inventions" for IP purposes, and thus subject to the much more limited extent of patent protection and for which each individual unit is an unprotected entity, are instead considered "new copies" of a copyrighted work, under which substantially more aggressive licensing restrictions are typical, and under which the definition of things like "infringement" and "derivative" are much more restrictive.

A software program should be an invention under patent law, and its copies should be considered discrete physical units. The code itself may be copyrighted, just as the plans for an invention may be, but the produced binaries probably should not be. Rather, they are units of the invention regulated as normal goods.

This has become even harder as we have gone to mostly-digital software distribution. You used to be able to easily resell software if you had a CD. Now you can't even do that (which, by the way, is no small motive for the companies; it completely obliterates the secondary market for software).


All software can be reduced to mathematical formulae, which cannot be patented, so no. There should be no such thing as a software patent.

I'm very sympathetic to the idea of disallowing software patents; to my layman's sensibilities, so many of them are ridiculous, or after the fact grants of something that has been done in practice for years, or predatory, or all that and more.

However, any physical machine or process can be reduced to the laws of physics, and you can't patent the laws of physics as far as I know. I don't think "reduced to mathematical formulae" is in itself a great argument.

Should the jacquard loom have been granted a patent?



The loom? Sure. The exact pattern on the cards themselves? Maybe not. I'd say no, but I recognize that US law currently says differently.

It seems like the old pattern of "Patent a certain method of weaving brocade", then later "Patent a certain method of weaving brocade...on (mechanical) computers!"

The exact pattern on the cards would seem to be a copyright issue more than a patent issue.

To the original objection that you can't patent math, I think some countries do disallow software patents, don't know if it's that reason or something else. But if they can disallow because math, shouldn't they disallow all patents because physics?

You're on to something here - physics vs math as a demarcation line is actually a thing:

German patent law for example has had the term "Technizität" as a neccessary property for a patent to be valid.


Money quote from the BGH (German Supreme Court): "Als patentierbar anzusehen ist eine Lehre zum planmäßigen Handeln unter Einsatz beherrschbarer Naturkräfte zur Erreichung eines kausal übersehbaren Erfolges." (GTrans:"To be considered patentable is a doctrine of planned action using controllable natural forces to achieve a causally overlookable success." :) )

i.e. You gotta use physics in your patent.

I have been coming back to this time and again, and today I still see it as the least shitty definition in regard to what should be patentable and what shouldn't.

Other national patent offices in Europe have used similar demarcation properties and it had served them quite well...until the degradation of patent standards started here as well around the turn of the millenium, with the first act being the EPO starting to issue software patents in rather blatant violation of its mandate.

No. Math is not physics, as physics is not chemistry.

Patents exist to promote novel applications of physics and chemistry, but do not forbid use of the fundamental laws, and should not forbid use of fundamental mathematical or algorithmic processes.

If you can violate a patent in your mind, it shouldn't be granted.

Not using a patent in a product or doing further research into the area, should reduce the time of patent on a logarithmic scale.

One could even claim that any mechanical device is ultimately reducible to a wavefunction, which is really just math.

When you can reproduce any mechanical device from its wave function, we'll talk.

When you can flip the bits required to operate a computer with pure abstract mathematics, we'll talk.

If there's a Theory of Everything then surely all physical existence can be expressed with mathematics.

In Europe at least maths can't be patented but applications of maths often can, just like software can't be patented as such but if it has real-world technical effect it can.

A new rocket launch trajectory may be "just maths" but if it reduces fuel consumption then it arguably isn't "just maths" and so will probably gain a patent.

This approach seems compatible with the quid pro quo intention of patents.

I mean, I'm very much opposed to over-broad software patents, but that's a fairly silly argument. Anything a human being produces can be reduced to mathematical formulae.

I think what they may've been referring to is something like the Curry-Howard correspondence, which describes the mapping between computer programs and mathematical proofs. Mathematical proofs can't be patented, and it could be taken to imply that (at least some) programs/algorithms shouldn't be eligible for patent protection.

> Mathematical proofs can't be patented,

Only because we've decided so. In the United States, for example, you can copyright the law.

> So you don't believe in software patents ?

There are two fundamental discussion schemes about software patents: (1) can we patent something that can be described as a mathematical concept and stuff. and (2) why would you want to patent software in the first place.

The first discussion is tricky. Comparing algorithms to mathematics, having patents that can be expressed through mechanical OR software, etc.

The second discussion is IMHO getting back to basics: Why do you want patent in the first place ? Why, in our free market universe, do you want to give to a restricted number of actors a monopoly ? Why not competition ? After all, if your neighbor decide to put a lemonade stand in front of your garden, and even if you didn't thought about this idea before, you are free to imitate/copy/replicate its business. Patents are inherently anti-free market.

To answer to this, you need to get back in time, and understand the reasons that led our ancestors to create patents: some technological inventions were extremely expensive, hard to copy, and companies were spending an unreasonable amount of money in them, say, to re-invent steam engine. This led to this bizarre exception, at the very condition that it would be limited in time, and really have a positive impact on the market (ie. NOT for the companies filling patents as sole incentive)

In software, there is no reason to allow patents. Don't get me wrong - a very very tiny fraction of software could be seen as needing patents (mpeg algorithms are often cited, probably engine injection algorithms too). But most of algorithms can be replicated by a bunch of engineers in an office. This is neither expensive, nor requiring specific provisions.

TL;DR: Software industry does not absolutely need patents, and because patents are very strong anti-competitive provisions, it should not be allowed in this industry unless there is a clear reason.

The big original reason patents came into existence was to reduce the number of trade secrets - an obvious block to progress; by ensuring that more inventions would enter the public domain (which was once rare and much tech was lost as a result, or very local.) See the history of forceps. Secondarily, as with copyright, you are rewarding the task of creativity. So, can you extend this first principle to software? Only if the software could operate in "a cloud" far from the user and therefore never be reverse engineered... which, oops, is exactly where we are. If there are no software patents, don't think that will mean that everyone gets to use the newest and greatest algorithms, it's more likely to mean that they will remain very private property that nobody else can build on top of. That situation could take us back centuries.

> will mean that everyone gets to use the newest and greatest algorithms

There are no such newest and greatest algorithms that are private and covered by patents. This is an illusion.

It's not just a matter of encryption or compression; there are always better and worse ways of doing things. We have seen large advances in the past there and elsewhere, you're predicting progress won't happen again - a bad bet.

Regardless of one's position on whether software patents should or should not be a thing, until you can convince Congress otherwise, they are a thing.

Do not underestimate the importance of software patents. It is wise to build a defensive patent portfolio.

I know this is no fun for people who are ideologically opposed to software patents, but as in most competitions, if the ruleset allows this advantage and you refuse to exploit it, you will lose to someone else who is less squeamish. There is absolutely a real lowest-common denominator effect that applies here.

If you want your startup to be acquired, you need something that a potential acquirer can't easily circumvent; that means a brand with user loyalty or intellectual property. Anything else in the software realm, Google et al could easily replicate it within a reasonable timeframe by assigning a small team to do so.

An artillery of your own ambiguous or general patents can also prevent lawsuits from others.

These are the rules of the game today, and if we want to play the game, we need to use this rule to our advantage. If we hate the rule, we need to find an effective way to petition Congress to change it. This usually means demonstrating how the current law is a clear risk for the only stakeholders Congress cares about: massive companies (particularly, major employers within their constituencies).

A lot of stuff that should be legally risky for big companies isn't, because the legal system is so convoluted and expensive, that they know only another well-capitalized company could successfully pursue a suit under such grounds. Since the big companies are generally advantaged by not "poking the bear" and instead accepting one another in a type of stasis, they avoid filing potential landmark cases that would dispense the legal ambiguity and strip the advantage against upstarts away from both parties.

"Don't hate the player; hate the game".

A bit like not believing in the law, something you can scream while they drag you to prison.

You should surely reject them as an acceptable practice, but it doesn't mean they can't present a legal problem. Some also say even if they reject them, they can still use them defensively.

But MP3 is what made it possible to share music on the Internet and to play music off of iPod-like music players. Shouldn't the inventors benefit from their work? Don't algorithmicists deserve the same protections that say electrical engineers do?

It's not mp3 that made it possible to share music on the internet but people who did not care for the business model of the music industry. By the time ipod were out mp3 was outdated and obsolete for a while but was still the most used format for several reasons (format of choice of the pirates, cheaper to make hardware decoder, lack of success of mp3pro and other proprietary successors, reluctancy to adopt free and open format).

No. They deserve the same protection as mathematicians.

No, apparently all software is created in free time, for hobby purposes. That's why all software should be free, open source and patent-free. Software devs can make an easy living off supporting their software.

Sarcasm doesn't do well on HN.

Decoding and encoding are separately licensed, right?

Why can my friends who studied medicine, physics, and mechanical engineering make money with their inventions, while I, software engineer, cannot?

Your friends who studied math cannot make money with their inventions either, and your inventions are really _a lot_ like math.

Since when is a patent required to make money off software?

That's not making money off software. That's making money off software AND business efforts.

And you want to make money off software AND patent efforts.

So, again, why do you think the patent is the driving factor to making money off software?

I've never seen any software worth patenting unlike with other disciplines, maybe it would help the case for software patents if they would be a need for it

I don't see why if someone invents a clever way to perform e.g. fluid dynamics simulations in software, that can't be patented just like in other disciplines.

Not all software is just shopping baskets, and website glue logic.

The relevant quote:

>On April 23, 2017, Technicolor's mp3 licensing program for certain mp3 related patents and software of Technicolor and Fraunhofer IIS has been terminated.

Does this mean that you can't get a license for MP3 anymore, or are they trying to say you don't need a license anymore? Are they opening it up, or are they shutting it down?

You don't need licensing for mp3 decoding anymore. All the patents have expired. Even Fedora ships mp3 decoder now.

There are still patents left for mp3 encoding though.

The last known decoding patent expired in 2015. That last known patent covering mp3 at all expired two weeks ago.

Edit: apparently there are two patents left in relations to encoder optimizations[0] not mentioned on the wikipedia pages. Those are probably irrelevant to modern encoders on modern computers.

[0]: https://madfileformatscience.garymcgath.com/2016/0/05/mp3pat...

Fedora wiki says:

> However, as of 2016-11-10, Fedora is now able to include MP3 decoding functionality. MP3 encoding functionality is still not permissible, because it requires patented technologies and the patent holder has not provided licenses that are compatible with Fedora's requirements.

So it looks like there are some patents relevant to modern encoders.

It probably hasn't been updated within the last two weeks...

Edit: The wiki page[0] links to the Fraunhofer page[1] saying that the licensing program has been terminated...

[0]: https://fedoraproject.org/wiki/Multimedia/MP3

[1]: http://www.mp3licensing.com/ which redirects to https://www.iis.fraunhofer.de/en/ff/amm/prod/audiocodec/audi...

I think you do not need one anymore. The patent were filled in 1997, that's now 20 years ago.

See: https://madfileformatscience.garymcgath.com/2016/0/05/mp3pat...

The last two patents seem to be related to an encode optimization and are not needed to write a decoder.

Which means it is possible now to include an MP3 decoder in GNU/Linux distros, etc. without paying the $2,500 fee.

I think 5,703,999 is about shifting between using joint and separated stereo for frames depending on the amount of difference between the channels to avoid cross-channel interference, this sounds like it might still be important for lower bitrate mp3s.

I don't think I know the technical details well enough to comment on 5,924,060, but I'll note that it has a priority date of 1986 and publication date of 1999... good riddance to submarine patents.

In any case, it is 17 years after publication or 20 years after filing (before 1995), meaning that both are expired.

So the Tunequest[0] article referenced got it wrong on the last two? I have a hard time figuring out what applies and does not apply to patents filed between 1995 and 1999

[0]: http://www.tunequest.org/a-big-list-of-mp3-patents/20070226/

The very last patent expired 16 April 2017, so there wouldn't be any point in licensing expired patents.

Some patents expired and no longer require license:



Looks like the parents have expired as of 2 weeks ago

Some number of years ago I wrote an mp3 decoder for a game I was working on. It allowed us to put hours of speech onto our 16Mb N64 cartridge (yep, it was quite a few years ago).

Knowing we had to get a mp3 license I contacted Fraunhofer who decided that a console game required a hardware decoding license which ran at $10,000s (if we had been a PC title it would have been a few $100 as I remember).

Legal got involved and decided they would license Rad Games Tool's Miles audio (since it already had a mp3 license and they could take any heat from Fraunhofer!).

Unfortunately for me legal then made me completely wipe the first implementation, carefully optimized N64 RSP vector code and all, and do a clean re-write of everything. We didn't use a single line of the Miles code, but they got a license fee and we had happy lawyers.

Did you consider Ogg Vorbis?

Given that the N64 was released in the North America in September 1996 and that the first stable release of Vorbis wasn't released until May 2000, it's very probable that OP couldn't have considered/used Vorbis.

It was probably 1999 that I wrote the first version. Ogg was still in development at that point. I know I considered a few codecs but mp3 was relatively well supported with 'reference' codecs.

Oh, ok. I thought "few years" was meaning actually few years. Did not know much about N64. After googling, it seems devs were still active in 2003-2005.

Mp3 was the only option that was relatively mature and had good enough compression ratios for what we needed (as I remember). Anything more computationally expensive was completely off the table.

Ogg and Vorbis hadn't yet discovered one another back in the N64 days, had they?

Perfect Dark?

Conker's Bad Fur Day.

I dont think Perfect Dark used that code, they used adpcm for their speech as I remember.

The reason I mentioned it is I distinctly remembered the MP3 copyright notice from the first screen of the game: https://i.warosu.org/data/vr/img/0015/50/1397802118061.png

Haha, I guess they did use my mp3 code - your memory is better than mine!

I see they have the Miles audio copyright in there too despite not only using it for the mp3 license.

Software falls under Process which is one of the four patentable subject matters for utility patents: Process, Machine, Manufacture and Composition of Matter. So software is patentable in the United States. However, software can also fall under the abstract idea exception and thus not be patentable.


Enfish LLC v Microsoft is the most important recent case. In its decision, the Supreme Court suggested that claims purporting to improve the functioning of the computer itself, or improving an existing technological process might not succumb to the abstract idea exception.


That's your dividing line. My guess is that Fraunhofer's MP3 patents improve a technological process and are valid.

So, now that the "but MP3 is patent-encumbered" consideration is finally dead-ish, what's the new balance on the debate between shipping {MP3, MP4-AAC, MKV-Vorbis, MKV-Opus} as music/SFX assets in software products?

Also, with the two sub-considerations:

• on mobile, where hardware decoding exists for only some codecs;

• on the web, where the audio APIs only accept some of these formats/codecs.

MP3 is the worst performer of the lot, but is suddenly more attractive because it's free and AAC is decidedly not. If you don't need compatibility (like for bgm/sfx) it's still a worse choice than Vorbis or Opus because it compresses worse. For sfx and looping music in particular, mp3 is also a mess because there's no standard way to play back samples with zero delay and zero extra samples at the end.

Mp3 is also really accessible. Anyone who's ever worked with audio in an amateur capacity knows how to create an mp3. I don't think many people would know how to properly encode an Opus file. I did an informal survey on Quora a while back, and I was surprised to see that almost every audio industry professional had no idea what Opus even was.

Mp3 is, for all intents and purposes, the animated gif of audio.

On the other hand, due to the encoder patent, most FOSS audio-editing tools ship without MP3-encoding support included.

My first instinct, on being asked to create an MP3 from a WAV + some effects, is to download Audacity, diddle around, and then export to MP3. But—and I always forget the first time I do it—you have to download an additional LAME plugin for Audacity to get that MP3-export to happen.

On the other hand, Opus is just right there in the Export menu; the codec is built into the app.

(On Macs, AAC is also "right there" in Audacity's export menu, due to OS-native Quicktime encoding support; but on other OSes AAC export requires similar fiddling to MP3 export.)

And, mind you, this isn't nearly as much of a concern on Linux, where usually the app package will have optional dependencies on the MP3/AAC encoder libraries, and so—if you have your "non-free" package repositories enabled—they'll get installed along with the software.

The patent situation really is like the animated gif but the technology involved puts it much closer to being the jpeg of audio.

> I was surprised to see that almost every audio industry professional had no idea what Opus even was.

Assuming you mean people who deal with audio production, I'm not surprised. Opus is primarily for real-time audio streaming, not for streaming music or for general audio compression or storage.


> Opus is unmatched for interactive speech and music transmission over the Internet, but is also intended for storage and streaming applications.

> Opus can handle a wide range of audio applications, including Voice over IP, videoconferencing, in-game chat, and even remote live music performances. It can scale from low bitrate narrowband speech to very high quality stereo music

Sure, their marketing copy says it's intended for storage, but that's not how it's actually used in the wild. In reality, it's primarily used for interactive/real-time streaming and not used for storage. I think it just boils down to the idea that the storage formats we're using are more than sufficient, and Opus doesn't really bring much new to the table there. It does, however, perform remarkably better than most/all of the alternatives when dealing with (e.g.) VoIP, so it's no surprise it's been adopted there.

True enough: the main type of engineer you'd expect to speak the words "Opus codec" is a telecom/SIP engineer, not an audio engineer. It always feels like tech with origins in telecom never gets much notice outside of that realm, even when it's much more widely applicable. (Look at the set of "core" Erlang libraries to see what I mean. Think most programmers have heard of even half of the relevant standards, even as they implement NIH competitors to those same standards?)

I vaguely recall that Vorbis requires a fair bit more processing to decompress, enough to be a potential issue 15 years ago, but on modern computers it's barely detectable and doesn't effect performance at all.

My recollection matches yours. I remember loading a custom firmware on my iPod, around 2005. Vorbis support was a big new feature; they'd just figured out how to get decode working fast enough on its 80MHz ARM7 CPU.

Probably Rockbox? iPods their clickwheel interface and it's software was amazing for the time, but other MP3 players generally had kinda icky interfaces. Rockbox fixed that :)

Yes, it was Rockbox. I followed it when they were first bringing it up on the 4th Gen iPod.

The "fast enough Vorbis" was likely the addition of Tremor, the first fixed-point Vorbis implementation. There's a nice (but with many old entries) chart of decode speeds here: https://www.rockbox.org/wiki/CodecPerformanceComparison

On newer hardware + software, like the Clip Zip, the gap is much smaller and the performance way higher than needed in any case (except for HE-AAC, the clear loser).

The problem now is battery.

MP3 almost always has hardware decode which is really power efficient. This isn't as bad nowadays since audio decode isn't that intensive relative to keeping your WiFi on.

opus can be decoded by hardware as well, (seek by name VP9 or webm) - most modern arm chipsets do so

...such as those used in the majority of Android/iOS devices? That'd be the important bit.

Wifi might be 100-200 milliwatts, but a laptop's speakers are likely to be 500-3000 milliwatts. Headphones are likely to be about the same as the wifi, but can vary wildly.

Bluetooth speakers/headphones change that dynamic slightly; the amp+required battery ends up on the receiving end, so you just have to figure what the Bluetooth stream itself costs.

I concur based on experience. I have a cheap music player that will play Ogg-Vorbis files but it skips and stutters. mp3 files play flawlessly.

I thought the "balance" was and continues to be that Vorbis delivers at least as good a quality as AAC at equivalent bitrates, if not better, better quality than MP3 at equivalent bitrates, and is FOSS which way tips the scale in its favor. I haven't seen and continue to not see any reason to MP3/MP4 in the manner you describe.

(And if Vorbis isn't on par w/ MP4, my impression is that it is certainly good enough at decent bitrates that it wouldn't matter.)

For new products there isn't really any room for debate anymore, Opus is the clear winner.

> For new products there isn't really any room for debate anymore, Opus is the clear winner.

If you mean "for low-latency live streaming/conferencing", I think you're probably right.

If you mean "by popularity", with a few exceptions all non-MP3 encoded audio you hear is (and has been for some time) AAC.

If you mean "by quality", Opus is comparable to AAC/HE-AAC audio quality, but Opus' Achilles' heel is hardware support.

Re: popularity; If you play PC or console games, you probably listen to a lot of Vorbis encoded audio.

Oh! Thank you. Q: Why not Opus instead of Vorbis for that use?

Because Opus was only released in 2012, so fewer people know about it.


It's the best codec in every dimension. It's royalty free, has low encoding delay, and has clearly superior compression efficiency.

> It's the best codec in every dimension

Battery life efficiency? MP3 is usually in hardware, these days. And battery life matters a lot.

(Not a gotcha--I have no idea.)

Better quality at same bitrate and way lower latency.

Observationally[1], most people never stopped using MP3s in the first place, so I imagine they'll just continue doing using them.

[1] Just looking through the sound effects and music asset folders in my sprawling Steam library.

I'm curious why you think shipping MP3 audio in software/games was ever seriously at issue due to patents? Decoding MP3s has always been non-problematic, and encoding MP3s was only ever problematic for a few years about two decades ago.

Some modern games use uncompressed audio because it frees up a few CPU cycles and the min/max balance on hard drives versus CPUs makes that choice desirable if you want to push the minimum spec. range on your game as low as possible (so that more people can play it). Though even then that's only necessary if your main game engine is hitting the CPU pretty hard to start with.

If it's assets in a program you're also supplying, Ogg Vorbis has been typical for a while now. Presumably Opus does better though.

I highly recommend "How Music Got Free" for an entertaining and illuminating read of MP3 and how it changed the world - for the better or worse.

It also delves into how Fraunhofer (and its creator) benefited monetarily from its adoption.

Link - https://www.amazon.com/How-Music-Got-Free-Obsession-ebook/dp...

I'm curious about what advantages and benefits, if any, the patents brought to Fraunhofer. Other than making mp3 slightly more difficult to work with especially on Linux, did it benefit them in any way? Did them benefit society?

Coming at this from an "this is a real-life example of software patents being {good|bad}".

Fraunhofer got paid lots of money from big companies buying patent licenses. Since they are a non-profit institution dedicated to scientific research, almost everything they make finds its way back into funding scientific research.

That said, we shouldn't generalize Fraunhofer's economics to the rest of the software industry. If it was just this kind of thing I would have zero problems with it. But we should be suspicious of software patents because the patent system implicitly rewards ideas that are broad (more things a patent touches, more $$$) and non-inventive (less work to come up with a patent, more $$$). Patents can be dangerous to human progress, and the patent system encourages patents that are. A broad patent could hold a whole industry hostage for 20 years, which is an eternity in technology. I sympathize more with folks who want to get rid of the whole thing for this reason.

The licensing money have probably helped the Fraunhofer Society do more research. Now whether that is the way you think research should be founded, and whether the net cost to society has been larger than the gain is another question.

For context, Fraunhofer Society is a massive public-benefit organisation aimed at applied research: while it gets some public funding, that public funding is given according to how much they make in industry contracts, which are the primary source of income. The idea behind this model is that they have room to do work that's not immediately profitable, but on the other hand have to keep applications in mind and work effectively to still get industry work.

Money, of course. All kind of companies (including megacorps) paid for it.

Even some private individuals like me (I think the company was called Fluendo?)

It made them a lot of money.

Were they involved in the creation of MP3?

Karlheinz Brandenburg at Fraunhofer invented MP3. (And AAC.)

(edit: not Stockhausen, thank you!)


Karlheinz Brandenburg... Stockhausen was a composer.

They developed MP3.

There's no real-life example of software patents being good.

Yes there are many, you just haven't seen any. Your perception is probably biased because you only get exposed to examples of bad patents cherry-picked by online media sources to drum up rage views. I've worked for a small company that got crushed by bigger incumbents but could only get a payout due to their software patents. However the much more common benefit of software patents are the countless startups and companies that got funded or had successful exits partially due to their patents. There are studies showing this sort of things. On mobile so can't find link, but see e.g. the work of professors like Robert P. Merges on ssrn.com.

I'm down on software patents too, but I'm not sure that's supportable without at least some sort of argument.

Given that the fundamental concept of software patents is bad, not only in principle but in practice, there's burden on anyone suggesting the good/bad question to validate that it's even a legitimate question. It's like the creationism/evolution "debate". I can simply say that there's no validity at all to the creationism arguments without providing any argument at all. Similar to the idea that I don't need to provide an argument to reject any other sort of baseless idea of false equivalence or unwarranted appeal to agnosticism.

If your premise is that "patents are not good" then why even ask the question about software patents?

It's rather disingenuous, the assumption with the structure of your question is "within the framework of 'patents can be good'".

Well, I indeed do not believe that patents are actually good, but there's far more merit to the unconvincing arguments that patents can be good than to any arguments that software patents can be good. There are good reasons that many people (particularly among those who understand software) fundamentally oppose the entire concept of software patents without opposing all patents.

"I'm obviously right, so there" is not an argument. If you want to convince a poor foolish Philistine like myself, you need to actually discuss your evidence and conclusions.

So is the FSF's campaign "Play Ogg!" [0] slowly becoming obsolete?

[0] https://www.fsf.org/campaigns/playogg/en/

Roughly in the way the "Burn All GIFs" [1] campaign became obsolete: the patents expired, removing the original objection, but nonetheless alternative formats are now better anyway.

[1] http://burnallgifs.org/archives/

PNG ended up being supported better than Vorbis/Opus though.

Sure, PNG instantly took over as the lossless non-animated format, but Animated PNG is still not widely supported (Firefox only IIRC?)

Mostly the "auto-playing animations without audio" stuff is moving to videos, e.g. https://developers.google.com/web/updates/2016/07/autoplay

Wow. I remember the initial hype for animated PNGs, meant to replace animated GIFs. We all expected it to happen in a matter of a few months, and that must have been nearly a decade ago now? It took 2-3 years before I simply gave up on checking to see if support had rolled out. To this day I have never seen an animated PNG in action.

Animated GIFs continue to reign as king. Image hosts' demand to reduce outbound bandwidth costs finally culminated in the .gifv "format". IIRC it's nothing more than a WebM or MP4 video without an audio track, targeted at the html5 <video> element.

What is the real reason why animated PNG was not rolled out years ago when we were all eagerly awaiting its arrival? Was the process encumbered for legal reasons, or was it simply not prioritized as an easily checked off box by the major players (ie: browsers)?

APNG was invented by Mozilla and not standardized with a "legit" organization (IETF/W3C). I guess Chrome/Safari/IE were reluctant to adopt a not-really-standard. Safari eventually did though, and now finally it's in Chrome beta.

Yeah "gifv" is what I mentioned, and it's the way forward. The bandwidth used by GIFs made from movie clips is ridiculous! APNG wouldn't be that much better since it's also lossless. But APNG will be nice for animations that have to be lossless.

APNG works fine here on Safari, and it seems to be supported by everyone but IE/Edge and (strangely) the new Opera w/ Blink, but that may be because Chrome apparently added support starting with 59 in March?

Easy test: do you see the ball bouncing here? https://en.wikipedia.org/wiki/APNG

Chrome too, as of Chrome 59 (currently in Beta): https://www.chromestatus.com/features/6691520493125632 and Safari: http://caniuse.com/#feat=apng

LOL, it was supported in the old Presto-based Operas as well :D One of the few "green to red" entries on caniuse

Like most of the FSF.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact