Commentary: Flash Is Dead Again (Yawn)
So, YouTube finally went HTML5 first with Flash Fallback, triggering another round of "Flash Is Dead" articles. Hey, didn't Flash die when Apple shipped the iPad without it? I remember reading lots of articles back then telling me it was so. Sure, it's a lot truer now, but don't send flowers to Adobe quite yet, as Flash will live on for at least several more years.
Let's review the current crop of "Flash Is Dead" articles. One of my favorites was in CNET; you can see the headline below. Stephen Shankland, who penned the article, has a very solid grasp of the codec and video standards space, though in this case, I think the headline (which authors seldom write) is a bit hyperbolic. Interestingly, he concludes, "although Flash Player remains widely used, that usage increasingly will be only on sites that aren't updated to keep up with current technology." CNET still uses Flash for their videos. Oops.
I found a similar inconsistency at Tech Radar, which doesn't deploy nearly as much video as CNET but runs Flash-based adverts on its site. Hmmm. Flash is dead, but we still need it to pay our bills?
Beyond these ironies, I ran a couple of quick surveys. First I checked broadcast and other premium sites who need to protect their content with DRM. To test, I visited the site using the latest version of Chrome, found the most accessible video, and right-clicked to see what technology they were using.
ABC was tough to figure out, because the right click didn't reveal any information. I scanned the page source code and found a reference to the Flash Player, but couldn't be sure they defaulted to Flash first. All the rest did though, including Amazon, though the site advised you to use Firefox or Safari with Silverlight for better quality.
Why are premium sites still using Flash? Because the Encrypted Media Extensions (EME) that enable DRM under HTML5 have more issues than a certain Patriots ball boy supposedly fingered for Deflategate. Click here to read a column explaining some of the issues with EME, and click here to watch a Streaming Media West presentation that covers the issues in depth. The bottom line is that HTML5 will ultimately work for DRM, but it will be a browser-by-browser, device-by-device situation. It's not going to happen overnight.
Moving on, I next checked Livestream and Ustream; both still used the Flash Player. Staying inside the streaming business, I checked the Brightcove website, which still used Flash, but noted that both Kaltura and Ooyala use HTML5 on their websites. Bravo, progress.
Just to be snippy, I checked the media customers Kaltura noted on their site. HBO and NBA TV still use Flash, though Paramount and Warner Bothers use HTML5 for free movie trailers (Warner via YouTube), and TMZ used HTML5. It's the same dynamic I saw on the first batch I surveyed—sites protecting their videos still use Flash; those that don't, use HTML5.
I next checked the top ten tech sites according to Webtoptenz. I found ome movement towards HTML5, but still a lot of Flash. Again, none of these sites care about protecting their video, so DRM was not an issue. In the end, I didn't do a ton of research, but as far as I can tell, I did more than any reporter who proclaimed that Flash was dead, and in several key markets, Flash is very much still breathing.
What's It All Mean?
What's the net/net? HTML5 is definitely coming, though the Media Source Extensions (MSE) that enable higher-end features like adaptive streaming and live streaming still aren't implemented in all browsers and devices. While YouTube has the programming budget to design around this, most smaller publishers don't.
For premium content publishers, DRM will be a huge gating factor, as you can tell from the first table. Ultra HD codecs will be another huge gating factor, as there is no free option for HEVC, and it's unlikely that the VP9 decode will ever make its way into Safari or Internet Explorer. How any publisher will universally deploy a UHD codec or codecs across browsers and devices is very unclear at this time, with the notable exception of Adobe Primetime, discussed below.
Then, of course, there's inertia; For every fire-breathing, flag-waving, sandal-wearing Flash hater, there are dozens of CFOs and CMOs and other C-level types who just don't give a flip: "Hey, if Flash is good enough for broadcast sites, it's good enough for us, Why spend the cash to changeover when we've got phablets to buy?" My 17-year-old daughter has an opinion on everything, let me tell you, but though she knows what Flash is, she doesn't care about it one way or the other. The bottom line is that most people who aren't technical just don't care, and neither do the folks who would have to pay the freight to change over from Flash to HTML5.
Said another way, beyond the technorati, the value proposition for HTML5-based video has never really been firmly articulated. So why would we expect everyone to fall into line just because YouTube finally changed from Flash first to HTML5 with Flash fallback? It really is nothing more than switching the order of a few lines of HTML code.
Finally, let's address the misguided "Flash Is Dead" comments. From where I sit, the simplest solution for both the DRM issues and UHD issues is Adobe Primetime, a premium, Flash-based platform that could keep Flash kicking for a long, long time. You can read about why in the same column referenced above.
To wrap up, YouTube delivers a tremendous percentage of video traffic, dwarfing all the other sites listed combined. But it's still just one site, with a very unique set of interests that are very different from those that deliver premium content, as well as many other sites. Drawing broad-based technology conclusions based upon YouTube's practices is just a mistake. At least in the premium content space, a site like ESPN or CNN is a much better canary in the coal mine than YouTube.
Overall, I don't disagree with Shankland's conclusions; I just think they're about 12-24 months premature, and don't take into account the potential impact of Primetime. To paraphrase Mark Twain, rumors of Flash's imminent demise are (once again) greatly exaggerated.
Everybody's using HTML5 and MPEG-DASH these days, right? Not so fast. Reports of Flash's death have been greatly exaggerated.