Youtube Ditches Flash, and it Hardly Matters

Youtube Ditches Flash, and it Hardly Matters

A year ago, the largest video site on the net ditching Flash would have been a blow for Internet freedom. Today, it's a bitter reminder of how the three big commercial browser vendors -- Apple, Microsoft and Google -- Netflix, the BBC, and the World Wide Web Consortium sold the whole Internet down the river. [...]

Now, you can access all of Youtube videos without having to use Adobe's proprietary software, so long as your browser supports the W3C's version of Adobe's proprietary software. If you're using Firefox, you can access all of Youtube's videos without Flash, except that in some cases, you'll need their version of the W3C-standardized "Encrypted Media Extension" -- which requires that you use proprietary software. From Adobe.

Meet the new boss, same as the old boss.

Previously, previously, previously, previously, previously, previously, previously, previously.

Tags: , , , ,

22 Responses:

  1. Nate says:

    Cory Doctorow is an inconsistent pedant. Yes, he hates DRM in any form. (It killed his father at Bến Tre? Invoking Vietnam is only one step away from Godwin.)

    Flash is not just about DRM. It's an entire VM, API set, etc etc. Youtube and other browser video players initially took advantage of the fact that it had decent built-in codecs and a major installed base. Real Media lost out (not that it was ever much in the running).

    Netflix came along much later and built DRM in ActionScript to run in Flash. It's a Turing-complete environment, so you can do stuff like that. Before that, Microsoft built DRM in native code via ActiveX. There was much suffering. Hackers came along and exploited Flash flaws from here until eternity.

    Getting rid of Flash is a net win for security. A lot of surface area that only stuck around for video playback is replaced with.... just video playback.

    Netflix/Google/Microsoft directed W3C to add DRM back as a separate HTML5 extension. If you want to play Netflix videos, your browser has to support it. But this is a much narrower use case that "Turing-complete environment".

    This is supported by the fact that Mozilla is able to sandbox the CDM component. They will supply an anonymized token to it to lock content to the device, but block it from any hard drive or system access. It gets encrypted frames in and supplies plaintext out. Any compromise of any Turing-complete VM implemented in the CDM (which is proprietary) will be contained.

    Youtube, which has designs on becoming its own movie studio, will start encrypting its "premium content". Then, you'll implement your Youtube downloader to sniff frames coming out of the Firefox sandbox and it will keep working as before. Besides, watermarking is the once and future king, catching users that repost the video.

    This is not the huge loss Cory is whining about. It's a small win (better security, more minimal API than "your entire computer and hard drive") but many things stay the same. But to religious zealots like Doctorow, anything other than complete eradication of DRM is a loss of some fundamental human condition.

    • nooj says:

      The internet: "Hey, I'm a great way to move files!"

      HTML: "Hey, I'm a great way to contextualize the file you want to move!"

      DRM: "Fuck you, see with your eyes, not your hands."

      • Nate says:

        That's funny, but I experienced it differently. It was "hey, these floppies are easy to copy" and then "Huh, I have to buy this one because the nibbler doesn't work. How did they do that?"

        I may be showing my "get off my lawn", but I wonder if there's an alternate history where consumer media never merged with PCs. That is, you'd have optical storage in your PC but the same thing didn't connect to your TV. The DRM would stay in the devices that played movies and music and the computer would remain a communication and computing device.

        Even today, nobody is raring to watch TV on their laptop. Ever see a big ole Dell on the seatback tray, blocking the headrest TV? That's all about "choice" and "cheap". That person doesn't want to pay the airline sky-high fees and/or they want to watch something other than what's being offered.

        In my alternate world, a startup cable company began kicking studio butt in 1990. You know, HBO + Comcast or whatever. Instead of their dark night of the Napster, everyone realized they had to compete on choice and cost and by the time your airline had screens, they were high quality and could access any movie anywhere.

        Meanwhile, the internet would thrive with fair use video/audio like jwz's Hellraiser edit partly because it would be hard to transfer large amounts of video from the thing in your living room to the thing talking to the Sum of all Human Knowledge.

        Also, Sept. 1993 never happened, ix.netcom.com failed to launch, and my laptop screen has been a black and white high contrast Retina resolution since 2000. In short, the creation of stuff via network + computers never merged with passive consumption. TVs stayed TVs and computers stayed computers and you can all get out of here because it's time for my liver pills.

        The End.

        • jwz says:

          In this alternate history are there also dirigibles and air pirates, and for some reason nobody thinks cats are cute or ever points a camera at one?

          • Nate says:

            Hey, I didn't say it was feasible. But it's fun to wonder if there may have been a point where consumer electronics and computers could have maintained their independence. I'm not talking about "connect your camcorder to your Amiga", which was bound to happen. I mean "DVD players are too expensive so I'm going to reuse the storage unit in my PC to play back movies on this uncomfortably heavy laptop."

          • Nate says:

            A small example that did happen is consumer DVD-R media isn't recordable at the location where the magic block of CSS keys goes. Therefore, you can't physically make bitwise duplicates without commercial grade equipment, dividing piracy applications from fair use and self-produced video.

        • Nick Lamb says:

          Your alternate world makes no sense because computation is the borg. Why would it choose to arbitrarily skip exactly one, very important aspect of human life?

          Nobody has ever "realised they had to compete on choice" because they don't, we have copyright, and so long as copyright exists in anything like its current form they never will. The closest we've come to forcing the issue is mass disobedience of copyright law, which is facilitated by the exact thing you say shouldn't have happened.

          Some time before this modern age of video being data and just a click away, an important 20th century movie adaptation of a literary work, Kubrick's "A Clockwork Orange" was literally not available in one of the first world democracies, the United Kingdom. Its copyright holder refused to allow it in cinemas, on VHS, or on television, so to see it legally you had to travel to another country. Copycat incidents (yes, really) caused Kubrick to decide that his movie shouldn't be available any more, and Warner Brothers as copyright holders simply withdrew it altogether.

          As to "nobody is raring to watch TV on their laptop" maybe so, but I literally haven't watched anything on an actual TV for some time now. When I watch stuff I do it on a PC or a tablet (I watched all of "The Great British Bakeoff" while sick in bed, courtesy of the BBC iPlayer). Friends who are into television as a passtime own a projector or huge panel display hooked up to a cheap PC running XMBC or its moral equivalent. I only wish that their freedom to watch absolutely anything without paying meant they had better taste. Do you know I've seen Nemesis 4? One of my "friends" broke the law in order to obtain a copy of Nemesis 4 and then showed it to people they knew. Why not just invite people round for an actual good movie, and then poison their snacks if you don't like them?

          • Nate says:

            Modern consumer electronics has always had computation in it (the original CD player in 1980 required a microprocessor). But I still don't believe that distribution of movies and music requires a union between a general-purpose computer (laptop, desktop, server, etc.) and consumer equipment.

            The reason why your friends go to all this effort to hook XBMC up to their projector is some combination of "choice" and "cost". You couldn't watch Kubrick and they didn't want to pay for "Nemesis 4" or something.

            In the alternate world, consumer electronics went on a frenzy of improvement in the 80's - 90's that beat the computer industry. The portable MP3 player followed the Walkman/Discman sooner, and the iTunes Store appeared in the late 90's just as the Internet was taking off. You could easily connect your home stereo to the Internet, your car syncs to your music jukebox collection, and there's no need for your computer to touch all that. It just works, most people pay money for what they use, etc.

            Of course this is a ludicrous vision. It would require so many companies to not be screw-ups that it never could have happened. It took massive piracy + Steve Jobs + iPod just to make the music industry sell their tracks properly online. Sony could never build a usable jukebox/MP3 server. The telco/cable monopolies still get in the way.

            The reason I care about this (a little too much) is that I founded a company in 2000 to provide the software layer to make this all possible. Think "Android model" but for digital audio. We'd license the player software to CE companies really cheap (if not free) based on a modified Linux base. We'd provide custom services on top of this software, such as streaming, P2P distribution (with payments), music locker for your CDs, etc. We'd take a cut of the revenue stream from those services, which could be white-labeled as the Sony Music Store or whatever.

            My focus was on usability of the consumer side -- you wouldn't have to know where your music was stored or anything. Sync between all your devices registered with a single profile was automatic, using free space wherever it could find it. If you streamed a track, we'd bump a least-played one out of flash and cache it locally.

            Anyway, we failed for a number of reasons but the biggest was that it would have taken massive funding to push our own CE device as well as major leverage with the music labels to get them to not sue you. Meanwhile, Napster was killing the market with "free" as the competitor.

            I don't bring up all this to wallow in the past, but to point out a hopeful potential future. We don't have to accept Google or Microsoft as overlords, and maybe the computer will retake its rightful place as a creation platform, leaving consumption back in the realm of CE devices.

            • Tim says:

              But I still don't believe that distribution of movies and music requires a union between a general-purpose computer (laptop, desktop, server, etc.) and consumer equipment.

              I was not aware that there was a received dogma that movies and music require such a union. That would be foolish as "consumer equipment" continues to exist.

              That's a cheap shot, but honestly you're not making a lot of sense in any of these posts. Like, what the hell does this even mean?

              The DRM would stay in the devices that played movies and music and the computer would remain a communication and computing device.

              In our real world, DRM is implemented in devices that play movies and music, some of which happen to be general purpose computers, and said computers continue to be communication and computing devices. What's so great about your world where, presumably, jackbooted Blackbeards parachuting from zeppelins slash the throats of anyone who tries to play back audiovisual media on a computing device? Why should anybody be nostalgic for the era of fewer options? Other than maybe Sony or BMG.

              Or you. It seems like this is really just you pining about how you coulda been a contenda. Sorry I guess? But I'm having a hard time seeing what exactly would be better for me if your vision had won over the industry and we had fewer options (and more DRM!) as a result.

              • Nate says:

                Sorry it's not clear. I'll have to write about this some time on my blog. It actually ended up being 3 companies over the past 15 years and 2 were successful.

                The summary is that mass media delivery (i.e., billion dollar movies and top albums) was totally screwed up by the CE companies letting PCs take the technology lead in the 90's. The resulting scramble led to DRM screwing up the general-purpose computing environment, when the two never needed to be so closely linked.

        • nooj says:

          Even today, nobody is raring to watch TV on their laptop.

          I think BitTorrent disagrees with you.

          • Nate says:

            You're confusing the power of "free" with the method of viewing. An iPad is a better portable TV screen than a laptop, and a TV is a better home viewing experience than a PC monitor.

            • nooj says:

              No, you're confusing the method of viewing with the control over content that it allows.

              If mass media delivery hadn't decided to fuck me to exponentially line their own pocketbooks, I wouldn't have gone to other sources of content. I still can't listen to even half an album on pandora. I can't build a playlist or crowdsource a playlist. It's 2015 and I can't netflix my music? So for you to say the RIAA, et al. was "totally screwed up" by PCs is a little silly, even if it's true.

              Also, my monitor is bigger than my TV, with far higher resolution. And my living room stereo has multiple inputs. Just cause it's on a laptop doesn't mean it has to suck.

  2. Nate says:

    Also, isn't it kind of odd to be watching 2 hour movies using a hypertext web browser? Shouldn't there be some multicast software package optimized for this purpose? I think this speaks more to the inability of software companies to fix their distribution problem. (Packaging/installation too heavyweight and unreliable? Force everyone to install a browser since only a couple companies are capable of creating/distributing those.)

    In my alternate world, software agents also exist and computing is truly distributed.

  3. His argument would have been stronger if he would have made it about Netflix instead of YouTube. AFAIK YouTube has never actually used any DRM, whereas Netflix uses DRM via Silverlight, and now supports EME DRM in some browsers.

    We (Mozilla) haven't shipped any EME DRM implementation yet that I'm aware of, which is why you still need Silverlight to watch Netflix in Firefox.

    Killing the Flash plugin is strictly a positive thing. It's a horrible mess in terms of security and stability. DRM is a pox, and I wish we weren't spending resources to implement EME, but I still think it's a net positive to have that vs. Flash. I also wish that either we had the market leverage to affect change here or that Google wasn't a sellout to the media overlords who want DRM, but unfortunately we live in the real world.

    • jwz says:

      If you don't think the bullshit countermeasures I'm constantly fighting in youtubedown don't count as DRM, you're splitting hairs. Just because it's not very good DRM doesn't mean it's not DRM.

      I had to write self-modifying Perl code with a Javascript parser in it.

      Self-modifying code, Carl.

      • I would not disagree with you, but isn't that tangential to his argument? That's just a symptom of giving them a Turing-complete language to work with. He seems upset about the black-box DRM blobs, not about web apps being able to run arbitrary JavaScript and obfuscate things. (God, if that sort of crap was all we were up against I would say we were in a pretty good place!)

        • jwz says:

          You said "Youtube has never used DRM", I say yes they have.

          They're just as evil and complicit as the rest.

          But anyway, eventually we're going to live in a world where programs like youtubedown and wget will need to rely on some library that lets you take an HTML5+Javascript roiling-pile-of-hell and run it in a sandbox or even full VM to see what comes out. I know such libraries exist now but they aren't common or easy to use (and more importantly, aren't pre-installed) but eventually, out of self defense, they will need to be. That will be a horrifying state of affairs, but on the plus side, it will finally make it possible to write screen savers in Javascript.

        • グレェ says:

          Ted, you and Nate earlier, seem to perhaps put too much weight into "Turing complete' languages and environments.

          Ever since Wolfram's 2,3 postulate was proven (http://www.wolframscience.com/prizes/tm23/solved.html), the notion that Turing complete is very meaningful beyond a programmatic sense has more or less eroded.

          At least, when you reframe it with the knowledge that a system about as complicated as three light switches and something to store memory in is also Turing Complete.

          (don't get me wrong, there were other simple implementations too, this one is just kind of almost impossibly reduced, if there is a smaller one [some have theorized a singular electron is Turing complete, and I can probably envision that as true, though whether it is meaningfully programmable is beyond my knowledge] I may just kind of just wonder if such lower level constructs even have development toolsets which are discernible or usable from a standard human reference to space time)

          Are people doing anything interesting with minimalistic Turing complete systems? I am unsure, I will say Tromp's BLC from the IOCCC in 2012 was a bit mind blowing: http://www.ioccc.org/2012/tromp/hint.html

          I mean, the fellow even used BLC to implement a brainfuck interpreter that was 1/4 the size of the previous smallest brainfuck interpreter. o_O

          Put another way: DNA is also Turing complete, and much faster than silicon; but we already knew that, yes some are using it programmatically already, and this is where things really start to get weird: http://www.extremetech.com/extreme/134672-harvard-cracks-dna-storage-crams-700-terabytes-of-data-into-a-single-gram

          So, while Turing completeness is a useful property, you may be more hard pressed to find a language or system which does not have the property of being Turing complete. Really, try to imagine one, does it do anything meaningful and it's still not Turing complete? That seems more far fetched, paper tapes of infinite length notwithstanding.

          Whether anyone meaningfully harnesses such minimalistic properties and their potentials, well, we still contend with Windows systems, and worse, OpenVMS, so I think we can already see that the answer to having the property of something which is Turing complete is not necessarily very meaningful, and more often very grim.

          Thankfully there's still js1k and the 4K and 64K demo scene competitions to at least showcase some artistic applications of minimalism. Indeed, those may be some of the best reductions of art even deterministically possible in an abstract intellectual sense. Personally at least, I find the real time demo scene to be the apogee of most programmatic prowess. If only it were more popular among start up culture, or web development. At least there's Steven Wittens and acko.net; but yikes. . . if you think about the underlying monstrosities of codebases (OS, then browser) layered to make such things happen; it still makes the minimalism of the C64 and Amiga demo scenes seem reductionist beyond compare; likely never to be repeated so much as imitated.

          But yay, Turing completeness!

          Sadly, we still have fax machines and other decrepit technologies which do not maximize such potentials. ;-/

          • jwz says:

            I just want everyone to know that this comment right here, this is why my blog allows comments.

            • nooj says:

              BTW, Stephen Wolfram gives a good talk! They have a big showcase at iSX, and he will demo a bunch of new features and probably pontificate on science and whatnot. Highly recommended, if there's a way to stream it.

      • Also I look forward to when youtubedown has to grow a full JavaScript interpreter in order to continue to function! Years ago I hacked basic CSS parsing support into wget in order to support fetching images specified in "background-image: url(foo)" rules and people immediately asked me about handling JavaScript. I passed on that.

  • Previously