DRM/EME in HTML5 - an American thing

I am happy to see that this last weekend there has been much discussion about the on-going DRM-discussions in W3C. One commentator argues that the W3C is anyway not very important and that therefore actors like Mozilla don't need to leverage any influence over their standard discussions. At Ars Technica it is written that the DRM-standard at W3C will not be worse than flashplayer, for instance, and that therefore it doesn't matter.

The idea seems to be that it doesn't mean any harm to standardize at W3C. I would argue, because of the W3C patent pools, that this is wrong. Anti-trust authorities in both the EU and the US are increasingly looking at anti-competitive uses of patent licenses and other licenses, because especially in high-end technologies like electronics and software it's leading to a lot of problems. In the EU, we additionally have the Microsoft browser case which establishes that a de facto standard can be equally abused as an adopted formal standard. The W3C standardisation process sets up a patent pool which is free for everyone in the world [was corrected on this point, but it makes an even more important difference for the anti-trust, I guessall the W3C members. A non-enforcement pact. That makes W3C very resilient to anti-trust investigations, which would not be the case if a DRM standard was adopted outside of the W3C as a de facto standard. If one does not believe that DRM is conducive to online innovation, and if one does not believe that Netflix is the pinnacle of God's creation, one may not want to provide them with that out.

Especially large commercial interests, including Google, Netflix and Microsoft are highly aware of this. Those who were irritated with Microsoft's behaviour in standard organisations in the late 1990s may want to seriously consider if it's more OK to do the same thing just because it's one decade later.

I've spent the last two weeks interfacing a lot with the Mozilla developer community. I got to attend the Brussels session on DRM on Sunday October 6 at #MozSummit, and I had a nice teleconference with a legal counsel from California on Friday October 11.

My argument is the same as the one I have raised at the ICT multi-stakeholder group at the European Commission June: technology can be political and we don't want to standardise behaviours and norms that we are not socially and politically ready or willing to accept. For me the open web and the internet are tools that we use to express democratic values, political discourse, parody, use and sharing of culture - DRM does not fit in this image because it is made not to do that.

For instance Cory Doctorow has raised the argument of DRM additionally stopping innovation: it seems to me that when one particular business model gets codified in a standard, that will lead to problems for entrepreneurs who do things differently. This is an extension of the problems we have with copyrights and neighbouring, where the legislator decided many years ago on a particular business model which has turned out not to be so useful as time passes.

But hearing the Californian Mozilla people, and also reading the Ars Technica article, I am starting to wonder if this is not actually also a problem between the US and the EU: the Mozilla people from California were not aware that Netflix does not provide services in all of the European Union. They are afraid of losing the Netflix customer base in the US if they don't implement the EME standard, and if they don't they feel that Google, Microsoft and Apple anyway will. However, in Europe Netflix is only launched in a highly limited number of member states. In the countries where Netflix has launched, normally they have strong competition from domestic streaming services.

In the US there is no strong distinction (as I understand it) between broadcasting rights and copyrights or neighbouring rights. In Europe, these three rights are not only legally distinct but also separate per member state. So Netflix, which normally is not in possession of copyrights for the works they stream, instead get covered by broadcasting rights. But the copyrights to be licensed, and the resulting broadcasting rights, will be legally unique per member state and in the European Union particularly the member state specific broadcasting rights are covered by special terms that linger on our markets since the television cross-border liberalisation of the early 1990s. In Europe one is allowed, as a commercial and as a non-commercial consumer, to transfer a broadcasting right of someone else from one country to another. In the EU, therefore, there may be legal changes required before Netflix can use the EME standard in the envisaged way. Netflix appears to have started to lobby the Swedish government - other EU member states I don't know. It's only relevant in this case to understand that Sweden is not the European Union, but one out of 28 member states all of whom have different legislations in this area.

There are additional problems with standardising here: the standard will spread to other media, like newspapers, digital libraries, audio files, and so forth. This is obvious. The browser will turn from being an open space where we can find and easily use information, to a place where actually end-users end up being quite controlled, especially by large commercial interests. This is largely unconducive to ongoing efforts by public broadcasters, newspapers, archives and educational institutions to create more open access to the cultural heritage of the European Union, but there would be similar setbacks in the US.

We are organising a discussion on the issue of DRM in HTML5 in the European Parliament on Tuesday October 15th (in two days) at CET 11:00-13:00. Alas, we've had many last minute cancellations, primarily from those who are involved in making the DRM standard which shows just how contentious this issue is. We will be trying to organize good streaming, and a recording of the conversation for later view and this will be tweeted through the @exile6e account. For us, it is important to distinguish between the role of a legislator and democratic governance and CEO Jeff of the W3C. Ultimately, I was elected to represent and make decisions in accordance with the general interest of the European public. The W3C does not carry this legitimacy.

9 kommentarer

Actually Europe has to gain more institutional weight. I can't see why the EU does not spent the 30 Million to finance their own browser engine in Europe, or convince Apache Foundation to get a European branch. We need European weight on the technology side and can buy influence quite cheaply: Invest in bottleneck technology development. Keep in mind that in the EU there is now a simplified recognition process for web standards in public procurement and there certainly these questions have to be asked and will be asked

Mozilla Foundation has a revenue of 160 Mio$, that is peanuts for the state. Europe should buy them out and migrate their entire operations to Europe, get us full access to technological decision making. And webkit? Well, some research project or state aid for a EU based Euroweb foundation, and we hire 50 developers that implement the ideas that provide added value for European values.

For the W3C we have to ensure dominant participation by Europeans and a relocation of its services back to Europe. We cannot accept an NSA dominated network infrastructure.

@Andre: Nope. Government-funded software projects like that don't work.

Typically, the people who are supposed to make the software will release some piece of crap (or nothing at all), and disappear with the money.

At best, after several years, you'll get some sub-par browser that's less good than Firefox, Chrome and IE, and nobody will use it.

Either way, it's a complete waste of money.

EME's DRM component is called a CDM. A CDM that works with netflix is required to interoperate with Netflix offering.

Neither the W3C nor Netflix nor Microsoft nor Google nor Widevine nor any of the Widevine subsidiaries nor any Operating system is currently offering a Netflix compatible CDM, Documentation and samples to Open Source software.

Mozilla could reverse engineer what IE11 and Chromebook uses, but this would be disallowed because of the DRM-circumvention clause in the DMCA.

This means that only proprietary browser like IE11, or Chrome builds from google (which are proprietary, with bits tacked on like video codecs under license, and a CDM) can legally/technically support EME.

This means that open source/community browsers like Firefox, Chromium, Webkit builds, Community Android browsers and so forth will not be legally/technically be able to support a CDM that works with Netflix. They might be able to implement their own CDM, but Netflix will not support that.

This is a grave threat to the web as Open Source/community browsers are the backbone of innovation for the web. If the web reverts to a state where only proprietary programs can legally/technically serve the full web experience, then innovation on the web as a whole (not just in a few novel niches) will cease.

@Andre, for someone reading this blog, you sound awfully like a clueless executive from some big corp. Mozilla is not just some company, it's a non-profit that many people join despite the lower salaries because they believe in it; furthermore, it's an organizer of a big community of volunteers. You could buy the name and trademarks, but you can't buy the trust and passion of all those people.

You can't build a successful community project by dumping money into it. We've seen the failures of Nepomuk and other EU-sponsored projects. If the EU wants to help, they should instead offer to find Mozilla as it is, and give them independence from Google.

There isn't a DRM spec at the W3C. There is EME spec - DRM *extensions*, which is an interface for running DRM plug-ins.

A browser that implements 100% of the W3C EME spec will *not* be able to play Netflix at all. W3C EME is only one part of the two-part system. To play Netflix browser will need to implement W3C EME spec *and* license non-standard, closed-source fully proprietary DRM plug-in code from Netflix (which is what Google did for ChromeOS).

Therefore W3C's patent protection is useless, because it only protects the nearly-irrelevant easy bit.

Technically it's very similar to HTML4 [object] tag and Flash plug-in. A browser that supports [object] can't play Flash without Flash license, and the fact that [object] is part of the W3C HTML4 spec doesn't help much.

Then, if I understand correctly, could Mozilla just support EME and the users would simply install the Netflix DRM plugin, the same way as they installed Flash plugin? I completely agree on the fact that this is bad for the open web and all, but my question is just on the technical aspect: would using Netflix with Firefox be any more difficult than using YouTube with Firefox (pre-HTML5)?

You could probably more usefully direct such inquiries towards @hadleybeeman or @hsivonen. They both work with Mozilla and should be able to respond how the Mozilla implementation of EME is undergoing.


I can personally reflect that @klewr is "bending the words" a bit. Of course we can relable every three-letter-combination as another three-letter-combination whenever the first becomes unsexy, but ultimately both of the three-letter-combinations amount to the same loss of control for the end-user. In fact, the latter incarnation of the content control three-letter-combination (EME) might even be worse for the user than its previous incarnation, since it relies on compliance and cooperation of the hardware of the user, as well as the software.

Lägg till ny kommentar