“Everybody knows” vs “Let’s test it!”

“It isn’t what we don’t know that gives us trouble, it’s what we know that ain’t so.”Will Rogers

Sean Christmann of EffectiveUI benchmarks Flash and HTML drawing and video on various current devices. Flash drawing performance is usually a multiple of each of the “HTML5″ engines, save for video where the availability of hardware acceleration controls all. (Jan Ozer had more on video acceleration last year.)

Background: Rich-media performance is hard. Even simple audio-mixing is hard, when you figure in latency and switching… HTML5 audio problems today are reminiscent of the first cross-OS media runtimes in the mid-1990s. Realworld video isn’t as simple as just adding a VIDEO tag… you need to make it work. Most runtime engines can’t afford to go as deeply into optimization issues as those who write engines for a wider base. Even when one runtime increases its JavaScript performance or its drawing performance, that doesn’t help when you need to run on more than one runtime.

Will the HTML runtimes continue to improve? Of course. Will Flash continue to improve and further diversify its support? Of course. Will commercial social-media accounts assert that Flash is slow and a battery hog, despite evidence to the contrary? That is, of course, possible. But here’s how Sean wrapped up his testing:

“The Flash VM performs really well on mobile chipsets and I don’t see any evidence here to support the idea that Flash is slow on smartphones and tablets.”