Flash Into (Not so) Thin AIR
Adobe Flash Player for Mobile is dead. Long live Adobe Flash video playback on mobile devices.
This is the message that Adobe will be trying to spread over the next few days, as it faces a slew of "told you so" reporting in the technology press.
We called into question Adobe's Flash Player for Mobile strategy a few months ago, weeks before other sites picked up on the issue. After the demise of the highly anticipated HP WebOS-based TouchPad, the veracity of the multi-platform mobile strategy Adobe announced at the 2010 Adobe MAX conference looked increasingly suspect.
The demise of the TouchPad and subsequent deep-sixing of WebOS, plus the underperforming of the Research In Motion (RIM) Blackberry PlayBook with its QNX operating system, left Adobe with only one viable mobile platform on which it delivered Flash Player—Google's Android OS. So we expected to hear Adobe issue a modified strategy at the 2011 Adobe MAX conference held last month in Los Angeles.
Instead, there was an almost eerie silence around video as a whole in the 2011 MAX keynotes, replaced by an emphasis on two other hallmarks of Flash Player: gaming and heavy interactivity. Turns out that Adobe was choosing its battles as carefully as it was choosing its statements of joint commitment to HTML5 and Flash.
After Adobe's decision to kill off Flash Player for Mobile (and curtail additional work on home media devices such as set-top boxes) the strategy is becoming clearer: Where it can compete with HTML5—in areas like gaming and heavy interactivity—Adobe will put resources and efforts into staying ahead of the looming groundswell of HTML5. Layoffs aside, one expects to see Adobe redouble its efforts in these areas, with the help of embedded AIR capabilities that were touted at the 2011 MAX show.
In areas where it can't compete with HTML5, or where there is less of a distinction to be made between Adobe Flash Player plug-in capabilities in the browser and those of the browser playback of non-Flash video files, the company will continue to emphasize its role in championing HTML 5. Part of this strategy is the consistent refrain that Adobe Flash Media Server is able to deliver video to a variety of mobile devices—including Apple's proprietary HTTP Live Streaming for iOS devices—that were never going to support the Flash Player plug in.
This point was driven home in a blog post by Pritham Shetty earlier today. Shetty is Adobe's vice president for video solutions, and he emphasized both what Adobe will support and what it won't support.
The former includes the desktop Flash Player plug-in and mobile applications, via AIR embedded in mobile applications. This doesn't mean that Flash video playback on mobile devices is dead; in fact, killing off the web browser plug-in capabilities means that Flash video playback will live on in a much more controlled environment.
After release of Flash Player for Mobile 11.1, Adobe says it will no longer put resources into either that plug-in or the ability to use Flash in the web browser on an internet-connected television. At first glance, this abandonment of mobile Flash Player capability seems odd, as Adobe alluded at the 2011 MAX to having solved the problem of allowing Flash video playback on every Android device.
In that same MAX keynote, however, Adobe didn't spell out how it had solved the problem, instead shifting to talk of embedded AIR within applications. That shift was intentional, as we've all heard over the past two days.
The shift not only relieves Adobe of the need to continuously update support for all Android devices—past, present, and future—but also shifts the burden of video playback support to the application developer, since all developers now have the power of embedding Flash Player playback capability into their own applications—via embedded AIR—and testing the app in a controlled environment before it's put out into the wild.
Adobe follows Microsoft on this path, as Microsoft has issued a porting kit for its Smooth Streaming client so that Smooth Streaming can be used on any device, if the device manufacturer chooses to implement the port. The Adobe strategy for home appliances and internet-connected televisions will be via the Adobe Flash Player Porting Kit that allows Flash-enabled web browsing on home devices.
Adobe is also allowing Research in Motion, as a source code licensee, to continue forward with enhancing Flash Player for Mobile on RIM's QNX operating system. It's another smart move on Adobe's part, especially since RIM's PlayBook sales have fallen off. Without the need to address PlayBook, Adobe only needed to solve the problem of Android.
Adobe had a huge burden in supporting all the variations of Android OS, as testing done by my consulting firm Transitions earlier this year showed a staggering number of combinations of WebKit, RTSP and operating system kernels. Hopes for consistent video playback were pinned on Adobe's plug-in, which we covered in a Streaming Media Europe session on the topic of video playback on Android devices, noting that many of the browser-operating system combinations in Android rendered video playback on the native operating system inoperable.
The problem was so severe that the curve of keeping up with all possible Android combinations would have required Adobe to throw so many resources at variations of Flash Player for Mobile that the group could've inadvertently become the largest resource drain within the company, ironic when the problem lies with Google's allowance of free-forking in the Android OS rather than Adobe's Flash Player for Mobile plug-in architecture.
The one major concern for mobile application developers in all this, though, has to be app bloat. If there's a requirement to embed AIR in every single app, the amount of storage required on a single device that uses multiple Flash-based media applications could mushroom uneccessarily.
App bloat will invariably cause two things to happen. First, the number of iOS applications that exceed the 20MB limit for downloading on mobile networks will increase, forcing the user to download the applications when they are on a Wi-Fi network. Second, and perhaps more importantly, the search will begin for a way to download or embed a common run-time platform on mobile devices, which may lead to a more accelerated demise of Flash usage on these devices, led by the device manufacturers. Either scenario may lead back around to a need for a plug-in architecture in the next year, as HTML5 still isn't ready for prime time scalability.
Or maybe the move to eliminate Flash Player for mobile devices will lead toward MPEG DASH. Shetty's blog post says the company wants to "focus on unsolved problems, like standardized dynamic streaming and DRM for HTML5, that are holding back reach and monetization across devices" and those are exactly the things that MPEG DASH is trying to solve, if DASH 264 is taken into account.
A common file format for HTTP video delivery might be just around the corner, according to panelists at Streaming Media West