HTML5 and DRM
kevinwhitman.com
kevinwhitman.com
1. It's easy for a consumer to say DRM is evil or to claim that it doesn't work anyway. It's a lot harder to literally bank on those things by irrevocably giving away the content your start-up has spent much time and money building.
2. It's easy to assume that any protection will be cracked eventually, and for mass market content like Hollywood movies and pop music it probably will. But many smaller companies are both producing niche content for smaller audiences, making them rather less likely to be cracked, and aiming for rather more modest profits, meaning that if someone rips their stuff and puts it somewhere Google can find there is a real danger of catastrophic actual losses.
3. It's easy to say that HTML5 video is the future and Flash needs to die. Apple and Google would be more credible when making such claims if their own browsers weren't riddled with bugs and limitations in this area and if they gave straight answers about exactly which video formats they support.
4. Open Source software that doesn't support locked up content or formats will give an inferior user experience for the foreseeable future. Whether that loss is acceptable to the users and developers of such software is something only they can decide.
Just to be clear: I am not a fan of excessive copyright protections, certainly not a fan of DRM, generally in favour of openness, and sincerely hoping that we can find a viable model for our business that fits with those beliefs. I should also add that I'm not writing on behalf of anyone but myself in a personal capacity here.
But from a realistic point of view as a businessman, those arguing for unrestricted HTML5 video vs. DRM/Flash need a far stronger case than they typically make today before their arguments are going to be as convincing as they would like.
From a commercial point of view, DRM "works" if it ultimately prevents more lost sales than it causes.
In a small niche market, where plenty of users are quite capable of right-click-saving a video but relatively few would know how to break more serious DRM, it is entirely plausible that such DRM would prevent a lot of casual copying.
As I said, I don't personally like this idea (and in fact our test programme isn't using DRM right now) but there have been lively discussions about this issue and the truth is that no-one really knows what would happen either way. I'm more interested in empirical evidence than thought experiments, assumptions/generalisations that small media will work the same as Big Media, or the wishful thinking of people reading on-line forums who have the luxury of risking nothing when they downvote a situation they don't like.
Edit: The other thing to keep in mind is which of these decisions are one-way. If we decide to launch with DRM and it turns out not to be effective for us, it costs us relatively little if we decide to back it out and serve everything unlocked later. On the other hand, if DRM is effective in our market but we serve everything unlocked initially and people take advantage, Pandora's box is open.
I think it's different when you're talking about permanent content someone is "buying" vs. temporary content someone is only paying to access as pay-per-view/rental/all-you-can-eat subscription, though. In the latter case, the pricing model is going to be set at a level commensurate with the usage model.
That can be a good thing, if it means more people get to enjoy more content than they otherwise would. For example, maybe someone who only tends to watch films once would not want to pay full price to buy every DVD but would be happy to have a flat-rate monthly subscription to an on-line streaming service. They can only do that if someone manages to run a viable flat-rate monthly subscription service, and if there's no enforcement, it could become a one-sided deal where the consumer takes full access but only pays the restricted access price.
This might not be a problem if the users in your market are basically honest, decent people and you can still make a reasonable profit while overlooking the occasional indiscretion. This is the outcome I'm kind of hoping for personally for our own business.
On the other hand, it's a big risk to take in a world where right now piracy is rampant and much of it is casual copying. I suspect that over time, as the first generation who have had always-on everything for their entire lives grow up, we are going to see quite a few cultural changes in terms of how we use the Internet. I think one of the changes will be in our attitude to sharing and the business models we use to produce and distribute content: a lot of things are already moving to supply content as part of a flat-rate, all-you-can-eat service rather than a series of one-off purchased products. Again, that might be no bad thing for either the producers or the consumers in the long run, but until that becomes a social norm, assuming it does, I can certainly understand businesses that want to make sure both sides are keeping up their end of the deal.
I know of at least one spotify competitor using this exact setup to fit both the mandatory DRM'd streams from major labels and still allow streaming to flash / silverlight barren environments.
For the big media the main goal of DRM is control.
They want to dictate what you cannot do: they want to segment the market by devices and countries, they want to force different deals with different distributors, they want ability to add unskippable ads, they want media played in players that report who is watching what and when.
Even if piracy didn't exist, studios would still insist on DRM to get all the other "features" it provides.
2. Less likely to be cracked? That's wishful thinking at best. It's already been shown countless times that smaller companies content will be cracked just as readily as larger companies. It's not as if they are using some other form of DRM.
3. Which bugs afflict you in the video playing department? I've found that playing videos without Flash is fairly simple. Granted, you need to provide 2 videos, but you'll what you need. Are you referring to lack of DRM support in these browsers? In fact, I'm not really sure what you are referring to here.
4. That's not true. Rather, content that includes extras like DRM will provide an inferior user experience. I'm sorry, but even Valve, who I love, doesn't have as great an experience as I can get if they'd remove the DRM. And it's because of the DRM they use. If you attach DRM to your content, you are degrading the UX. There is no argument against that. It's simply the way it is. Even if you simply put your content in a membership area, I have to log in rather than simply see that content. If I have to install some 3rd party software to read that content, it's a degraded experience. I'm sorry, but it's the content providers that choose to degrade the experience. At least man up and take ownership.
Finally, you claim people like me need a far stronger argument. I guess that's fair. We could all use stronger arguments. But at the same time, the arguments for DRM aren't strong either. The only argument in favor of DRM that I can get behind is: "It's my content, I can do what I want with it." Yes, you can, and I support that. Beyond that, not much more can be said.
2. If you've got any serious data to share on this point, I'd appreciate links. I've been researching the real effects of copyright, infringement, DRM, etc. for a long time, and I've found few studies with anything resembling a sound methodology and unbiased reporting, and even fewer that consider anything but mass-market Big Media operations.
3. I'm not sure this is a good place to enumerate every likely browser bug we're currently tracking, but I'll give a few examples:
- Some browsers are not detecting the end of a video properly (no JS event fires/can't replay -- not really much doubt that this is a bug).
- Some smartphones play videos fullscreen whether you like it or not (unhelpful if there are other controls beyond the universal video ones that users should be able to reach with one touch -- this is an active choice, but limits the quality of the experience we can offer to users).
- Several mobile devices reportedly support H.264 at a certain profile and level, but in practice seem to require more specific settings (which don't always seem to be clearly documented in the guidelines published by the browser/OS developers -- this may or may not be down to bugs, but without clear specs and completely standardised formats it's in any case a significant drain on resources just investigating and testing across a wide field of devices).
I'm not referring to DRM at all here, just the problems of using HTML5 technology that isn't sufficiently standardised yet relative to the one-size-fits-all-except-iOS of Flash.
4. In our tests, H.264 in an MP4 file (using the features that are supported in browsers) gets modestly better quality at a given bit rate than a WebM stack and completely outclasses Ogg/Theora/Vorbis. If you have a browser like Firefox that for cultural/legal/whatever reasons won't support H.264, then the best result is that Firefox users will see lower-quality video or need more bandwidth to see the same quality. There's not really anything ambiguous about this.
As for DRM, if it ever interferes noticeably with legitimate users, that's obviously a bad thing. Again, my personal line is that any technology we do decide to use to protect our content in any way should be transparent to someone using our service legitimately. But if we're talking about someone visiting a web page to watch a video, why does the user care what technology is being used to achieve that as long as it looks and sounds as it should? If in some alternate reality I could use a plug-in on iOS smartphones to provide a better user interface instead of relying on Apple's compulsory full-screen presentation, or if we decided to produce an iPhone app for the same reason, would you still object to requiring 3rd party software so a customer could use our service more easily?
> But at the same time, the arguments for DRM aren't strong either.
The argument for some form of transparent DRM in our internal discussions is simple and compelling: either it works and we potentially make more sales to legal customers at the expense of pirates, or it doesn't work and we're probably no worse off than we would have been anyway. Unless either the implementation costs are high or we're at risk of losing significant numbers of legitimate customers for some reason, it's essentially a plan with no drawbacks other than the bad taste it leaves in the mouth.
Once again, I feel I should stress that I'm writing personally here and not in any official capacity. I should also be clear that I am playing devil's advocate to some extent, as are other people when we debate this in-house. No-one is evangelising DRM here, we're just not buying the anti-DRM arguments without reading the full brochure either.
> The only argument in favor of DRM that I can get behind is: "It's my content, I can do what I want with it."
Well, yes, and to the extent that "I can do what I want with it" means "I can run a viable business that helps my paying customers" and "People who don't pay us for it don't get to have it" I don't think anyone here has much of a problem with that.
If genuine users have even the slightest problem because of any technological content protection measures, that's a serious issue for me.
If it costs us a lot of time and/or money to implement those measures and they don't generate a worthwhile return, that's also a serious issue.
However, being nice to people who are blatantly trying to rip us off rather than paying a price that, frankly, almost anyone in our target market can easily afford is... not a priority, let's say.
This is not entirely true. I think you mean to say we will have access to less content, which is true. (And a fine trade-off for me.) My user experience with copyrighted content is actually quite nice, everything works as expected.
I'm curious what the trade-offs in the other direction are. I tried coming up with a few:
1. Your customers will face more friction getting your content. This will cost some customers. This also means more customer support issues.
2. DRM technology will be expensive to develop/license and maintain. Especially if you have content that needs to be used across platforms. Distribution is also trickier (do users need custom software, etc).
3. Smaller customer base. Some platforms will inevitably be excluded and some people won't use DRM stuff.
I never watch series episodes twice, and I'd rather pay small money to see them once than to buy DRM-free video files. Statistically, I must be the minority because Apple stopped offering series rentals. :(
(And kicking the two divs which hold that warning out of the way with Firefox 10's web inspector and a "display: none" gives a perfectly reasonable page. Why not just show that all the time?)
The latter is of course nonstandard, but cooperative browser vendors could easily implement the pure SSL authentication. I can't imagine Flash or Silverlight video is any more secure than that. There's of course no way a truly open source browser would be trusted in this way; the certificate would have to be hard to extract.
(This comment might make me sound pro-DRM. I'm not, just making some technical observations)
Adoption of such DRM would be at mercy of PHBs from big media, so they wouldn't choose a vendor that values openness and end-user freedoms.
In that case just use Flash or Silverlight, as "HTML5 DRM" won't be any better: you won't see it in FOSS browsers (due to licensing/goals), Apple devices (Apple has their own DRM already and market power to keep it that way) or niche platforms that aren't "commercially-viable".
P.S.: Bing Cache works
There are the bandwidth-divining problems since you'll want a higher quality stream on a fast connection than a slow one and HTML5 video has no built-in way to do that. On iOS you can use HTTP Live Streaming via Safari, but if you depend on that you're locked into a single platform.
What is driving your startup to use HTML5 video? Are you trying to avoid licensing fees for DRM technologies? A lot of these problems can be overcome if you build an app, perhaps it's possible to use a hybrid approach in PhoneGap so you can maintain a HTML UI but play video via a proprietary video library that supports DRM.
[Bing Cache](http://cc.bingj.com/cache.aspx?d=691260178405&w=1e1573b4...)