If the EU finds royalty-free patent ceasefire clauses to be anti-competitive, then the entire foundation of standards-based interoperability for the web is at stake. These schemes underlie everything—even something as simple as addEventListener('touchstart'). And the entire realm of software encoding schemes has been poisoned by patent royalty schemes for far too long. I cannot think of anything worse for the state of media encoding on the web than regulator intervention to maintain that status quo.
“ The Commission has information that AOM and its members may be imposing licensing terms (mandatory royalty-free cross licensing) on innovators that were not a part of AOM at the time of the creation of the AV1 technical, but whose patents are deemed essential to (its) technical specifications," the paper said.
This concern is certainly valid, i. e. it is logically consistent. I would tend to agree with the sentiment that standardization is net-positive, but I know next to nothing about the commercial aspects happening behind the scenes of these processes.
Sounds like there are AV1 patents that are not free and are not interested in a cross licensing. Anyone knows what patents or parties are this?
In principle such a thing is problematic when the heavyweights get together and de-facto force others to sell at a low price. It is seen unfair as the big one finance this squeeze with profits from other markets where they but not the niche player gain.
I honestly can't understand if you are for or against this:
>If the EU finds royalty-free patent ceasefire clauses to be anti-competitive, then the entire foundation of standards-based interoperability for the web is at stake
hmm, that sounds against
Followed by
>the entire realm of software encoding schemes has been poisoned by patent royalty schemes for far too long
well, that sounds for
but then
>I cannot think of anything worse for the state of media encoding on the web than regulator intervention to maintain that status quo
uh, against if they maintain status quo but for if they say it needs to change? Don't regulatory agencies normally look into things because they think they might need to say stuff needs to change?
on edit: changed ugh to uh, as ugh is generally used as expression of disgust, whereas I was going for expression of confusion.
yes, those things sound like against - like I said - but this
>the entire realm of software encoding schemes has been poisoned by patent royalty schemes for far too long
sounds like it's for, right? If it has been poisoned by patent royalty schemes, and regulators are looking at patent royalty schemes here they must be for?
Also
> I cannot think of anything worse for the state of media encoding on the web than regulator intervention to maintain that status quo
>I cannot think of anything worse
is against, but
>regulator intervention to maintain that status quo
I mean, regulator intervention generally happens upending status quo, because when regulator intervention happens it means you will regulate something that was not regulated and thus what was status quo under unregulated system will have to adjust, even though of course big companies often improve standing from regulator intervention because they have the resources to handle the regulations and after a period of transition may be doing even better than before it still is the case that generally regulator intervention upends the status quo.
on edit: I mean tone is also against, so I normally just go by that, just some of the words don't seem to be as against as I would think from the tone.
I think you need to go back and re-read the article, or maybe find some other explainer about the state of media encoding licensing, because you're completely misunderstanding both what AV1 and the EU are doing here. The EU is intervening to protect "patent royalty schemes" because Google/Apple/etc are offering their patents royalty-free (that is, for free). Offering your patents for free is seen as anti-competitive, because the existing media-encoding-patent-licensing industry depends exclusively on royalties from patent trolling to stay alive. So because AV1 is a new codec without patent encumbrance that is offered for free to everyone, it risks disrupting the status quo, and killing the media-encoding-patent-licensing industry. Does that help clarify things?
I was confused reading the article, seeing your comment makes me at ease that I read the article correctly. I too don’t get why the EU needs to intervene when they are making things better.
Because others are loosing money and lobbied enough people to get the antitrust ball rolling.
No government serves the people. But especially not the EU. That is always the facade. This is an act of war against free standards.
Take German weather reporting for example. The tax financed weather service had a free app with weather warnings and weather forecasts. Without any tracking or advertising.
The commercial weather services sued it and where granted the ruling that forecasting of weather data payed by the taxpayers offered for free is forbidden as it hinders the commercial interests of the competition.
So either one pays for the app now or just has weather warnings in it.
Antitrust for the win (from the POV of the commercial entities).
> forecasting of weather data payed by the taxpayers offered for free is forbidden
That surprises me. The data itself is open (https://www.dwd.de/EN/ourservices/opendata/opendata.html: “According to a change in the Deutscher Wetterdienst Act, which has come into effect on 25.7.17, the DWD has been given the legal mandate to make its weather and climate information available mostly free of charge”), and trivially retrieved from their server, for example from https://opendata.dwd.de/weather/text_forecasts/
I don’t remember the URL, but they also have a freely available GIS server that serves weather warnings that is easily used as a layer on top of OSM data.
I don’t know the details, but I think the issue is that raw data is free, but writing apps on top of it is left to the market. I don’t see any reason why there can’t be a free open source app that uses this data.
The problem is not that *nobody* can offer weather apps for free, the problem is that the court has ruled that it is anticompetitive for the German government to offer a free weather app as a public service. You can read more details about the ruling here: https://www.dw.com/en/court-outlaws-german-weather-services-...
This means that any FOSS or free-as-in-beer weather apps must be funded out of a developer's own pocket, rather than as a public good.
The general legal principle that a taxpayer-funded service may not be in direct competition with commercial services is pretty much long-standing EU tradition and not a "war against free standards", see e.g. the requirement for public broadcasters to take down content [1].
The US follows a different tradition, wherein all works created by government employees are in the public duty [2].
Personally, I believe both ways of traditions are reasonable.
There are some policies that the EU introduce which make some level of sense but have unfortunate consequences.
I remember one such case with regards to Estonian air.
For those that don't know, estonian air was losing a _lot_ of money, but the Estonian government kept bailing them out. In fact it was considered the "cost of doing business" and took it as a form of investment in tourism.
Of course, this is still a private company, and subsidising the airline means that nobody can possibly compete at that price point.
So the EU ruling is sound, but the effect is that people are having a lot harder time getting to Estonia now.. which is a net loss for Estonia overall.
Saying the EU doesn't serve the people is a bit of a lie, but there are circumstances where the EU does something and a subset of people are worse off, even if it's a better idea.
the problem was favoring that one company over all the others. Estonia is free to subsidize flights to Estonia, but they should do it through tax rebate or something not through propping up one unprofitable state owned company.
The problems trying to develop a royalty-free video standard at ISO might be instructive here. MPEG floated a standard called IVC (Internet Video Coding) that was specifically designed to only use Option 1[0] patented inventions. After it was published they got a bunch of Option 2[1] declarations that effectively killed the standard. ISO doesn't actually require patent holders to declare what patents are being infringed if they are making an Option 2 offer, so there's no means to actually redesign the spec to not use those patents.
I'd like to introduce you to a man named Leonardo Chiariglione - co-founder of MPEG, if not the entire concept of FRAND patent pools. He has a rather long blog post from 2018 detailing ISO's growing patent problems here: https://blog.chiariglione.org/a-crisis-the-causes-and-a-solu...
The ending of this story is that Leonardo was fired or forced out about two years ago, and MPEG is dead - or, more accurately, cut into like five different pieces. There's a bunch of patent holders who are angry that patent revenues have fallen since the MPEG-2 days, angry that H.264 has zero-cost licensing for certain industries, and are deliberately trying to break the whole FRAND licensing system with conflicting H.265 patent pools to try and extort more money out of codec implementers. Leonardo tried to make ISO aware of this and several unnamed national entities vetoed him.
My conspiratorial guess is that one of those national entities is the EU, or perhaps ECMA[2]; being lobbied on by the same companies that gave us Velos Media and/or Access Advance. Those same people are probably whispering sweet nothings in the EU regulators' ears. "Look, Google is trying to use their market power to not pay for video codec patents![3] Ignore all the other sketchy stuff we're doing on the other end."
[0] ISO-speak for patents whose owners pledge to license them on a royalty-free basis
[1] ISO-speak for patents whose owners pledge to join a patent pool
[2] Can you tell I have limited knowledge of how ISO is organized?
[3] Leonardo himself actually also believes this and is arguably more opposed to AOM than the EU regulators are.
H.265 requires three patent license fees, since two other competing patent pools promised a handful of patent owners a deal on cross-licensing if they left MPEG-LA.
AV1 may not require a patent license, because Google designs the codec to only use unpatented or royalty-free patented inventions. However, there's still considerations of submarine patents - i.e. an unrelated entity that invented the same technology before Google. Parallel reinvention is not a defense against patent infringement - there is no inverse to prior art. This is the same reason why Microsoft refused to support VP8/WebM early on in IE - patent pools were supposed to provide an incentive against that sort of trickery. Then Velos Media and Access Advance came along and broke this, which is why the naysayers started piling on into AOM.
> "The Commission has information that AOM and its members may be imposing licensing terms (mandatory royalty-free cross licensing) on innovators that were not a part of AOM at the time of the creation of the AV1 technical, but whose patents are deemed essential to (its) technical specifications," read the questionnaire.
Cross licensing is the whole point for creating a free (as in beer) standard body.
There are too many existing patent in this field.
If mandatory cross-licening is seen as a problem, it will be impossible to create new video standard.
The problem seems to be “but whose patents are deemed essential to (its) technical specifications” phrase.
I read that as an investigation on this (real or imaginary; that’s what the investigation is about) stream of events:
- AOM members created a standard for which they, together, thought owned all patents
- they found out that isn’t true. Other parties hold patent(s) that are essential for the standard.
- they strong-arm the holder(s) of those patent(s) to join AOM and, with it, license their patent royalty-free to the others in the group.
Whether it happened I wouldn’t know, but step 3 should, IMO, be illegal. Royalty-free cross licensing is more beneficial for those who have the capital to build millions of chips implementing the standard than for smaller players. It’s like a few elephants forcing a mouse to have dinner together and splitting the bill.
My reading of that statement is that there's no problem with mandatory cross-licensing. Instead it's about other companies - who did not help with AV1 development - being strong-armed into joining the cross-licensing agreement.
Whether AV1 is patent free isn’t clear cut, it seems:
FTA: “…AOM and its members may be imposing licensing terms (mandatory royalty-free cross licensing) on innovators that were not a part of AOM at the time of the creation of the AV1 technical, but whose patents are deemed essential to (its) technical specifications," the paper said.”
Standard anti-patent language to neuter the patents against AV1 users and implementors. Until patent reform occurs, it's the best we got, and the closest we can get to keeping FOSS patent-free.
This is no different than the GPLv3 anti-patent language, imo.
Although I understand your sentiment, Mozilla was blackmailed by Google (using their yearly donation to keep Google Search the default); they were forced to either give up $300m/yr or kill Rust and cripple Firefox development.
Mitchell Baker, the current CEO and the CEO during this whole thing, makes $3m/yr, yet Firefox userbase keeps shrinking due to management throwing the engineering team under the bus.
I think it is safe to say that, although Mozilla isn't a shell company for Google, Mitchell Baker might be a shell CEO on Google's payroll.
I think at the core of this probe is strong-arming by market domination, hence antitrust.
In theory, patent cross-licensing agreements should be designed to pool up patents covering the standard, so that every user of such standard could use it regardless of patent portfolio.
In practice, with large enough market share to push standards on end users standard consortium can develop a standard inherently infringing on patents held by non-members and use market domination to strong-arm non-members into either reducing presence while litigating or joining the cross-licensing agreement, effectively rendering patents held unenforceable.
From what it reads that is the issue: patent pools coupled with market dominance used to expand said patent pool.
Other commenters mention an orthogonal issue - cross-licensed patent pools as a tool to effectively eliminate software patents. However, this misses the main point raised by this probe - which companies emerge victorious and in this case (such cases) the founding members emerge on top due to pre-established market dominance.
"It is also querying companies on AV1's compatibility with other video codec standards, such as whether a smartphone or a browser can support both AV1/VP9 and competing codecs HEVC and VVC simultaneously."
"And if AV1 were to turn into a new market standard, the commission asked, what would be the impact on companies’ ability to compete with AV1’s AOM licensing conditions, and their incentives to work on alternative video codec technology that may be non-royalty free?"
"Brussels is trying to understand the rationale of companies that didn't choose AOM and instead have chosen to license the AV1 and VP9 standard-essential patents through Sisvel's pool."
To be fair, as much as I dislike it, the commission may have a point. Basically AOM is using its size to make itself unavoidable and force free cross-licensing on what used to be non-members. This is mostly advantageous to the founders who know avoid paying fees to license the tech. From afar, it does indeed seem to hinder competition.
Patents hinder competition; this is a sensible response to take patents out of the picture. It's advantageous to anyone making use of media and looking to avoid the gatekeeping associated with MPEG formats.
At one point, copyleft was under fire on antitrust grounds, based on similar arguments. Fortunately that was defeated on the basis that it wasn't actually bad for consumers, quite the opposite in fact. Hopefully this attack will be defeated similarly.
I don't know every patents are anti-competitive. Generally I against software patent because some ideas are seems to obvious, but some algorithm/encoding technique could be worth patented (or say inventor should be rewarded), though it's pain to pay for each.
The only things patents achieve are to give huge companies the rights to squeeze small competitors out of the market, and to let patent trolls siphon money off of companies which actually make stuff. They need to go. This isn't limited to software patents.
Patents don't protect the little guy. They certainly don't help the individual inventor, even though protecting the inventor is always the given rationale.
It would encourage competition by allowing other actors to create alternative solutions and get money from it.
Competition can be good for consumer if new format technically superiors can emerge. Might be harder if this market is locked by huge companies that don't want to pay and push an inferior but artificially free solution.
That's an argument, not my opinion, I don't know what to think about this issue.
> Competition can be good for consumer if new format technically superiors can emerge.
Industry has had its chance and they choose to nickel and dime everyone to the degree that noone who uses HEVC can be sure they're not infringing someone's patent even if they've payed up every single patent pool.
AOM is others response to this ridiculous status quo, within the same pate t framework.
So if you don't hate the game, don't hate the new players, either.
Because all the members of AOM are posting insane profits including on their media part and media codec is an insignificant part of a smartphone price. Honestly the whole thing looks mostly like a play from big players to pay less on the encoding side. It’s mostly irrelevant to consumers.
Smartphones are hardly the only thing that cares about codecs. So does anyone embedding a video online, anyone building video editing software, any browser engine, any video call system, anyone building tools that play with video (e.g. AI/ML and every random "make you look like a cat" system), and a thousand other things that want to innovate around video.
...is the argument that these companies that didn't participate in the development of AV1 now can't avoid using AV1, and thus are forced to license essential AV1 patents under the terms offered?
(so, is the EU also saying that once a standard becomes big enough, every essential patent must be licenseable under terms the EU finds acceptable, even if the patent holder didn't agree to that beforehand or possibly even participate?)
It is not cross-licensing agreement per se being investigated, but
1. the inherent absence of some patents required for the standard/implementation to function in initial patent pool
2. use of pre-established market dominance of founding members to limit non-member patent holders' ability to create competing standard and effectively strong-arming them to join.
A very similar issue has been raised against Microsoft in the antitrust case regarding iexplore. Iexplore has been deemed "good enough" default that market dominance of Windows with iexplore preinstalled as the default effectively strong-armed everyone involved to at least maintain compatibility with iexplore.
> the inherent absence of some patents required for the standard/implementation to function in initial patent pool
I guess this is the disconnect; AOMedia worked very hard during standardization to avoid infringing existing patents held by non-members. There is a very strong belief that Sisvel’s claims on AV1 are fundamentally invalid, and while that would be nice to prove in court, Sisvel’s members arent stupid enough to directly sue anyone that can seriously fight back. Thus, the terms to license the actually essential patents to at least limit the damage Sisvel can do.
Anyway VVC is already technologically superior for anyone that wants a competing standard. If it fails, that’s 100% on the behavior of HEVC patent holders, not because of AOMedia.
The issue is more than they can’t avoid using AV1 and are therefore forced to license their patents for free to the rest of the consortium while AV1 probably infringes.
The comment you replied to literally said that they can't avoid AV1. Other commentors here have called AV1 an "open standard".
Note that the EU isn't claiming that the standard itself is bad. They're saying the cross-licensing terms are anti-competitive. The cross-licensing may have made sense for major companies like Apple, Google and Meta but it significantly disadvantages smaller companies which may need to rely on their patent arsenal as a deterrent but also face economic pressure to adopt the standard.
So which company is large enough to have patents that supposedly are essential to AV1 despite not participating in the process, is small enough that AOMedia overlooked those patents during the standardization process, yet is large enough to rely on those patents as a deterrent against AOMedia licensees, and obviously can’t be an NPE since they feel pressure to ship AV1?
But if these companies hold patents that are being infringed by AV1 they don't have to join AOM they could also… sue them.
This whole thing seems to be a breeding ground for submarine patents, where in the end nobody has an actual applicable patent to show but the thread of pulling one out of the hat is there to make things worse.
They have to join AOM if they want to use AV1 and they need to use AV1 if they want to be relevant because all the big players banded together and made that unavoidable. That’s the heart of the anticompetitive argument.
Most companies don’t exist to sue for patent infringement. They are not patent trolls. The point of competition law is that they should be able to do business without having to go to court with a conglomerate.
What does Fraunhaufer have to do with it? If we're randomly guessing which company might have instigated this investigation, I'd guess Sisvel, because it's easy to imagine for them to have an actual case: Sisvel has created an AV1 patent pool and iirc even made an open call for patent owners to approach them to get their patents added to the pool. It's easy to see how an anti-trust complaint could arise if someone left the pool or didn't join it because of real or imaginary pressure from AOMedia members.
What makes you think this is them getting tricked, and not just them looking after the interests of EU companies? The EU's role isn't primarily as a consumer rights advocacy group after all.
Fraunhofer is an institution in Germany positioned between universities and the industry. They provide applied research and license it to companies.
For example mp3 was mainly developed there. They hold many patents in the field of audio and video coding which obviously fund parts of the organization.
More recently, they established a new audio codec for Bluetooth LE under proprietary license which in my opinion does not have any advantages over OPUS.
At the same time not all of the criticism is fair because they are not a patent troll and anticompetitive in their license agreements.
Either the lack of technical literacy on the EU’s part is a real issue, or the EU wants non-European tech companies to fold up their Euro tents completely and stop selling their products in EU countries. That’s the logical end result of this and the gatekeeper nonsense.
Given the compromises the Apple is willing to make to keep their foot in China’s market, you can be sure that neither of the big tech companies is willing to give up on the EU market. Let alone allow the vacuum to be scooped up by a more favourable tech company.
Sisvel is pledging for 0.1~0.3 euros on AV1-capable hardware here, europe has done much worse (like with the OOXML standard) and unfortunately got away with it. This doesn't particularly affect the EU as a market, it's applied globally.
I doubt that EU wants companies to leave the market and they don't seem to encourage the growth of profitable European competitors to Google/Apple, so most likely it's the literacy issue and on top of that they're "surveying the grounds" to see how far can they go.
I mean, i would understand if the "gatekeeper" stuff was pushed by countries/organizations that have little to no manufacturing prowess, but it's Europe we're talking about and there's no way that the EU doesn't understand what they're doing (unless...).
This depends on the cross-licensing terms right? Like suppose AV1 becomes the standard format for web video, and as a result I use an AV1 compressor to create an AV1 video. Does that mean that if I ever come up with an invention in compression techniques and patent it I’ll then have to provide a royalty free license to AOMedia for this invention?
First, near as I can tell, they only claim you need a patent license if you distribute an encoder/decoder. MPEG-LA has claimed distributing encoded video needed a patent license as well, but backtracked on trying to charge for it.
Second, your agreement to license your own patents only covers Necessary Claims on the Implementation (both the specific implementation you're distributing and the reference.) Which by definition can't cover any future inventions, because if the implementation already infringed them then they aren't novel inventions. But it does explicitly claim to include IP you acquire in the future that applies.
Also, I think the intention is that the license you have to grant only covers implementations of AOMedia standards; if a VVC implementation or whatever also used your invention, that would probably need a separate license you could charge for.
> Also, I think the intention is that the license you have to grant only covers implementations of AOMedia standards; if a VVC implementation or whatever also used your invention, that would probably need a separate license you could charge for.
That is correct. Source: I helped draft the license.
Sure, and you can argue this is bad, but the US dosn't have a good record of policing anti-competitive behavior (especially in foreign markets for obvious reasons) so that heuristic probably works better than you make it sound.
Of course not everything large American tech companies do is bad, but if a large American tech company does something that seems even remotely fishy, it's safer to first assume that it's bad until proven otherwise.
EU citizen here. I trust the EU, my elected representatives, more than the adtech sleazebags and the patent cartels. I prefer to vote with my votes than with my wallet.
There's a long list of consumer-friendly EU regulations that I like. I'm also very supportive of EU privacy protections. I think the only problem with them is that they're not aggressive enough.
I completely agree with what the EU have done so far on a mostly positive front. But at the same time all governments and unions seem to regress to measures which lead to worse consumer protection and a worse security and privacy outcome in the long term.
I can't possibly support a union that wants technical censorship measures introduced while crowing that they are privacy focused.
And then there's the removal of the control point on the Apple App Store, which is admittedly bad for business but the best outcome for consumers as it allows them to control over billing in once place. The proposed outcome of opening the billing is corporate interest focused rather than customer focused and will only harm the latter.
And of course there's what they didn't bother to do. No one is perfect but some of the recent shifts are stupid and force an EU centric view on technology. This is the region that sold out all of its technical capacity (Philips, Nokia etc) without even blinking an eye on the regulation front and is now complaining when US tech companies step into the gap in the market they created. Hypocrisy is what it is.
I can't support a singular organisation which does so much good and so much stupid at the same time. It needs to be broken down into separate bodies as it's a monopoly itself.
They are following US pioneered anti-trust practices that went against IBM and Microsoft and allowed an ecosystem of diverse computers and browsers to develop. Limiting the exercise of power moves of monopolies is essential for innovation and landmark cases are vital to establish the boundaries of the rules. This motivates corporate lawyers to advise their clients accordingly before they do damage.
Markets where the main rule is pre-established power are in the long run destructive to society.
So a standard for interoperability developed with 'patent ceasefires' saying party X won't sue party Y if both share their patent portfolios with one another is considered to be anti-competitive behavior? In the US this is fairly common among tech companies that perfom this sharing under FRAND terms with one another, and saves a lot of headache in avoiding litigation.
The most popular existing codec AVC/H.264/MPEG-4 Part 10 is owned by the MPEGLA group with ~5500 active patents[0] in its portfolio, and the royalty sharing agreements[1] involved are free in small volumes yet become expensive at scale. A multi-tier fee structure for both hardware OEMs (think like the MPEG-2 license fee on the raspberry pi) and video distributors (Netflix et al) exist, where several large tech companies decided to band together in order to create a codec which isn't at the mercy of paying these license fees.
However, Sisvel wants to make their case against AOMedia's goal of a widely distributed free codec and extract up to 32 cents per device[2] based on their pool of a couple hundred patents[3] for both VP9 and AV1 codecs. We'll likely see this go to court and possibly drawn out over a long stretch of time if FRAND terms aren't able to be agreed upon between these parties.
Having a uniform royalty-free standard would be a net win for everyone involved- chip designers can add in an AV1 encoder/decoder without paying a license fee, software companies can freely encode & distribute AV1 videos, and end users have access to free video editing tools that wouldn't need to pay MPEGLA/Sisvel either. The alternative is waiting for the patents to expire and continue rewarding the rent-seeking behavior of NPEs/holding companies created to extract value at every possible step.
If this results is Sisvel's favor, I hope all Youtube stops serving VP9, AV1 or any newer codecs in Europe instead of paying up a single penny to rent seekers.
"Hi Europe, enjoy your low-res bitrate starved H264 stream, for eternity."
or for 30 years or so, after all HEVC patents expire and all of its rent seekers gave up on it.
Since the current US government is anti-tech as well, EU and CN governments are cranking up the pressure to get more money from the US tech. I guess this will last until next administration which will do the same to the EU companies. Once tariffs come for Bosch, and EU giants it will even out.
Both governments will get more money, customers will pay higher prices, tech giants will keep making money. Everyone wins, somehow.
I don't know what you mean by calling the "current US government" anti-tech.
During the Trump presidency the governing party constantly called for regulation of "Big Tech" and even quasi-nationalizing social media platforms over "first amendment" concerns (which don't apply to private companies).
By comparison the current US government seems fairly tame and anti-interventionist.
It's also strange you consider tariffs an equivalent of antitrust and consumer protection regulation. US tech wouldn't have to give the EU "more money" if it followed the laws and regulations of the market it wants to operate in. Considering how obstructionist US tech companies of any size have been when it comes to basics like data protection and privacy, I'd say the fines these companies have had to pay so far are actually too low.
As for whether the AV1 case is actually anti-competitive, that's for the investigation to determine. But using a market leader position to strongarm other companies into agreeing to terms that weaken their ability to defend themselves would definitely be an antitrust issue, if the investigation determines that is what is going on.
