Re: I strongly urge all supporters to reconsider the EME proposal. It is not in your best interests!

On Thu, May 16, 2013 at 8:33 AM, Zak Fenton <zak.fenton@gmail.com> wrote:

> >> I think you have mis-understood the point of the EME proposal. We are
> not here deciding whether DRM will be used on the web. DRM is used on the
> web and will continue to be used on the web for as long as it is a
> requirement of the licenses for content that web users want to watch. If
> you are opposed to DRM, it's that last part you need to change: either the
> licenses or what people want to watch.
>
> I don't believe I have misunderstood the point of the EME proposal
> (although I will be further scrutinising it over the coming weekend). It
> could be true that this would reduce the amount of "non-standard
> proprietary technology" required to implement DRM, but I personally don't
> believe it, I think this will probably end up as a big mess of "This site
> is best viewed on Windows 8" and "Sorry, this site cannot be viewed in your
> browser". In any case, I'm sure this will be great for DRM, it'll do
> wonders for DRM. But DRM is not a person, it does not have feelings and
> does not have a voice, and it is not nearly as important as the web's user
> base. What this proposal will damage is W3C's credibility as a standards
> organisation and the web in general. Browsers are big enough already. You
> should implement *your* DRM pipeline in *your* apps instead, because it is
> not welcome in *my* browser and has absolutely no place in *our* web
> standards. You may argue that I can choose not to watch DRM content or
> install third party DRM components, but this is actually the choice I have
> today. Tomorrow my choice might be between accessing the DRM-centric web,
> awash with a wide variety of untrusted installable components, and
> accessing only Wikipedia and torrent sites. In other words, the web is only
> now starting to recover from the damage products like Adobe Flash have done
> to the standardisation process, with popular sites like youtube finally
> starting to wean off the platform. Standardising the system of
> de-standardisation further than a simple embed tag, thereby lowering the
> barrier to entry for developers of non-standard components, will have many
> unforeseen consequences, not just on users but also on distributors,
> maintainers and content developers.
>

Please see my other mail about the intended role of browsers in determining
what CDMs are available. I agree that if we end up with "a wide variety of
untrusted third party DRM components" this will be a step backward. That is
not the intention. In fact that outcome is more likely if W3C does nothing.


>
> >> Avoiding a loss of quality - of one form or another - is actually much
> harder than you suggest. There are barriers - understanding and buying an
> HDPC ripper, re-encoding, a/v sync, redistribution, adaptive streaming etc.
> that make things far less convenient than sharing a URL to an unencrypted
> file (or the URL + key for an encrypted file)
>
> Far less convenient, sure. Impossible, no, not in theory or in practice.
>

Ok, so we agree. I claimed only "far less convenient". Noone claims
impossible.


> If this proposal does go through, then I'll prove it myself. No matter how
> convoluted the pipeline is, and how many proprietary components it passes
> through before exiting my PC through the HDMI port, the data at the end is
> still the same data. It is true that it's extremely difficult to record and
> encode it again and get a reasonable file size without losing any quality,
> but the slight loss in quality won't be anything new for pirate consumers,
> and is certainly avoidable in theory.
>
> >> Since when is pay-per-use or subscription-based instant on-demand
> access to content streamed over the Internet a "20th century business
> model" ?
>
> It's just a slight progression from cinema, popularised in the early 20th
> century (yes, there are cinemas that let you choose what to watch from an
> available media library). The basic difference between cinemas and modern
> pay-per-view streaming (apart from not being in a physical cinema) is that
> there is no longer any technical reason that the customer should not be
> able to watch that movie again, with the same quality, in their own time,
> without paying more. There is no need for the huge rolls of film and
> expensive equipment that made the pay-per-view model so successful last
> century.
>

I fail to see how the technological change makes a difference to the value
to consumers of the pay-per-view business model. Consumers have
always preferred pay-as-you-go type models for filmed entertainment, from
Cinema, to 'pay-by-watching-ads' broadcast, through VHS and DVD rentals and
now onto the Internet. Why do you think that the reduction in the marginal
cost of distribution and technical simplification in copying would make
consumers prefer a buy-to-own model ? Sure the distribution costs are
lower, so the price should be commensurately lower. When you pay for
content you are not paying only for the distribution medium. DVD rentals
never included the right to make a copy of the DVD to watch later, even
when that became technically quite easy.

Technology should provide platforms for as diverse a range of business
models as possible.


>
>
> >> The lone business model advocated by many opponents of DRM is a
> download-to-own one, in which the only thing which can be sold is full
> ownership rights in a copy of the content. This business model dates back
> to the invention the printing press, when copies or creative works could
> first be economically produced. Not that it's a bad model, it just isn't
> new.
>
> I didn't say pay-to-own was any newer, but since you disagree:
> pay-per-view dates back to the early days of prostitution and stripping.
> The earliest evidence of this dates to the 18th century BC, long before
> Shakespeare was charging per act. I'd argue that both models date back to
> very early human history in various forms.
>

Ok, so at least we can agree that there is nothing new here in terms of
business models.


> I'd also argue that pay-per-view is 100% unambiguously bad for consumers.
>

Why, then, do consumers consistently choose that option over buy-to-own for
video content ? Pay-per-view, and subscriptions models, enable the cost of
production to be marginalized across many small payments, rather than
asking for a single one-off payment from each consumer. Why should a
consumer who watches the content once be forced to pay the same as someone
who watches it 100 times ? Why shouldn't consumers be able to choose which
option is best for them: the large one-off payment or small incremental
payments ? Small incremental payments result in greater revenue - because
more people can afford them - which supports more content production and,
yes, more revenue means more profit, but we are not talking about charities
here.


> You might say it drives the creation of better content, but this is like
> saying that cheap third world labour drives prices down for consumers, in
> that it depends entirely on the companies doing it, and that those
> companies will simply do whatever they believe is most profitable to them.
>

This analogy makes no sense.


>
>
> >> DRM-protection of content on the Internet is and will continue to be
> widely implemented for as long as the licenses require it and the content
> is popular. You're in the wrong place if you want to change those two
> things. EME is about how the web fits into that reality and ultimately
> boils down to a rather narrow choice between arbitrary, bloated plugins and
> browser-controlled CDMs with consistent functionality and APIs (and, yes,
> before someone comments, CDMs share some properties with plugins, but the
> we think the differences are sufficient to be worthwhile).
>
> Correct, DRM is already widespread (enough) online. I'm concerned about
> users not understanding that DRM technology is in place, not understanding
> what DRM is.
>

So, placing control of the DRM under the browser can address that. Browsers
- that users trust - can give users appropriate information, warning and
control. In contrast, websites encouraging you to install an arbitrary
plugin might not be so motivated.


> I'm concerned about the prospect of "rendered them directly" mentioned in
> the proposal (presumably a typo for "rendering"), particularly about how
> cross-platform this will be and also how direct - for example, will "CDMs"
> be permitted greater access to my graphics card or operating system than a
> regular Javascript app, potentially exposing driver or hardware bugs to
> them which could be misused to gain control of my system? Will these "CDMs"
> be able to adapt to circumstances where hardware acceleration is not
> available or is not as effective (e.g. virtualised systems), if they render
> directly?
>

This case, where the protected media is rendered directly by the CDM, refer
to the case where the CDM is a platform capability - part of the operating
system - and the UA uses a platform API to access it. As part of the
operating system these CDMs probably do have privileged access to the
hardware.


> I can see that this could theoretically be more efficient, and perhaps
> even theoretically simpler and easier to manage, compared to current
> generation plugins implementing the DRM, but while you believe the
> differences are "sufficient to be worthwhile", I (and presumably a heck of
> a lot of other people) believe that there will be no *technical* difference
> in practice. This will simply be an official W3C endorsement of
> proprietary, non-standard, anti-consumer technology. For the average web
> user, this might not mean much - until this gains widespread use among a
> minority who are happy to pay-per-view
>

You'll find, if you look historically at DVD sales vs other
incrementally-funded options (DVD rental, broadcast etc.) that it's not a
minority.


> , and (being a standard and all) begins to find it's way into things that
> aren't intended to be pay-per-view, for example sites that have some
> commercial content and some free content may begin using the same codecs
> and encryption across the board in order to ease maintenance, thereby
> limiting user access to free content.
> In summary, I'm concerned that officially standardising this technology
> would lead to this technology becoming widespread in circumstances where it
> would not normally be considered at all.
>

That's a reasonable concern and I explained why I think that's an unlikely
outcome of this proposal in another mail.


> I do not have any opposition to the pay-per-view model itself (I won't buy
> it, but it's just another product), but I do oppose this model being
> encouraged as standard practice.
>

I don't think we should presume here to prejudge the decisions of consumers
in a free market. Internet pay-per-view and subscription video have been
universally welcomed by consumers in contrast to digital downloads which
are still far less successful.

...Mark

>
>
> On Thu, May 16, 2013 at 1:36 AM, Mark Watson <watsonm@netflix.com> wrote:
>
>> Zak, Piranna,
>>
>> I think you have mis-understood the point of the EME proposal. We are not
>> here deciding whether DRM will be used on the web. DRM is used on the web
>> and will continue to be used on the web for as long as it is a requirement
>> of the licenses for content that web users want to watch. If you are
>> opposed to DRM, it's that last part you need to change: either the licenses
>> or what people want to watch.
>>
>> The above is just the way things are. If there is popular content on the
>> web that is supported in one browser, the other browsers will want to
>> support it too and they will do wo with or without the W3C.
>>
>> What we are trying to do with EME is to *reduce* the amount of
>> non-standard proprietary technology needed to do this. I would reduce it to
>> zero if I could, I just don't know how to do that.
>>
>> See also
>> http://arstechnica.com/business/2013/05/drm-in-html5-is-a-victory-for-the-open-web-not-a-defeat/
>>
>> A couple more comments below.
>>
>> On Wed, May 15, 2013 at 7:43 AM, piranna@gmail.com <piranna@gmail.com>wrote:
>>
>>> +1, you've exposed very clearly your arguments and I totally agree with
>>> them.
>>> El 15/05/2013 16:37, "Zak Fenton" <zak.fenton@gmail.com> escribió:
>>>
>>> DRM simply does not belong on the web, it is contrary to freedom of
>>>> speech and it is of zero benefit to the consumers who fuel the web economy.
>>>> It will only make browsers and servers more complicated and more error
>>>> prone, restrict the ability of people to use the web, and waste CPU cycles
>>>> encrypting what is probably already widely available to pirates.. As any
>>>> technologically competent person is aware, unless you can stream the media
>>>> direct to the viewer's brain, there will ALWAYS be ways to circumvent these
>>>> methods: A paying subscriber to a channel or buyer of a movie can simply
>>>> record their screen and audio output (without any quality loss if they're
>>>> smart), freely sharing the result with others.
>>>>
>>>
>> Avoiding a loss of quality - of one form or another - is actually much
>> harder than you suggest. There are barriers - understanding and buying an
>> HDPC ripper, re-encoding, a/v sync, redistribution, adaptive streaming etc.
>> that make things far less convenient than sharing a URL to an unencrypted
>> file (or the URL + key for an encrypted file)
>>
>>
>>>  You cannot beat piracy with technology. Suffice to say pirates have
>>>> access to better technology, because they get it free! The only thing that
>>>> will slow the continual increase in piracy is better content, content which
>>>> is actually worth paying for, and better content developers, content
>>>> developers who people actually want to pay.
>>>>
>>>
>> Yes, and we are finding in market after market that people are very
>> willing to pay a modest monthly sum for access to range of content that is
>> available only with 'DRM-required' licences.
>>
>>
>>>  This proposal will not help anybody, it will only make web standards
>>>> more complicated, harder to correctly implement, and less reliable as a
>>>> result. I'm really beginning to lose my faith in standards bodies like this
>>>> to develop standards which are actually of benefit to humanity, rather than
>>>> standards which have been set by investors desperately trying to squeeze
>>>> profit from a 20th century business model. This simply does not make any
>>>> sense.
>>>>
>>>
>> Since when is pay-per-use or subscription-based instant on-demand access
>> to content streamed over the Internet a "20th century business model" ? I
>> spent just under half of my adult life (so far) in the 20th century and I
>> didn't notice any paid-for internet video streaming.
>>
>> The lone business model advocated by many opponents of DRM is a
>> download-to-own one, in which the only thing which can be sold is full
>> ownership rights in a copy of the content. This business model dates back
>> to the invention the printing press, when copies or creative works could
>> first be economically produced. Not that it's a bad model, it just isn't
>> new.
>>
>>
>>>  Older generations developed the technology, but it was my generation
>>>> that made the internet and the web a popular success. Without the freedoms
>>>> we had, future generations will simply move towards underground protocols
>>>> and networks that protect their freedom, creating a new safe haven for real
>>>> criminals. If this proposal is accepted and widely implemented, it will
>>>> perhaps mark the beginning of the end for the relevance of web standards,
>>>> but certainly not for freedom online.
>>>>
>>>
>> DRM-protection of content on the Internet is and will continue to be
>> widely implemented for as long as the licenses require it and the content
>> is popular. You're in the wrong place if you want to change those two
>> things. EME is about how the web fits into that reality and ultimately
>> boils down to a rather narrow choice between arbitrary, bloated plugins and
>> browser-controlled CDMs with consistent functionality and APIs (and, yes,
>> before someone comments, CDMs share some properties with plugins, but the
>> we think the differences are sufficient to be worthwhile).
>>
>> ...Mark
>>
>>
>>
>>>
>>>> Again, I strongly urge all involved parties to reconsider their support
>>>> for this proposal.
>>>> Yours sincerely,
>>>> Zak Fenton.
>>>>
>>>
>>
>