News for hackers. Stuff that matters.

A comment on Not So Fast On That HTML5 Canvas vs. Flash Demo.

Re:Not So Fast On That HTML5 Canvas vs. Flash Demo by beej71 on Monday September 20, 2010 @11:44PM (#1711385)
The original sources were pretty much apples-to-apples comparisons. They weren't fast, by any means, but the Flash and HTML really both did the same thing: clear the screen, draw the circle, move the circle. No dirty regions, just pure pixel pushing. Very inefficient, but generally equally inefficient.

Now to make HTML look better (er, rather make Flash look worse), Gruber keeps linking to the optimized versions of the HTML app.

I'm no HTML-hater. I love the tech, and have even done HTML canvas-related work-for-hire. (And plenty of Flash stuff for-hire, as well.) But I also love being informed, and Gruber is only giving half the story here.

Will someone optimize that Flash piece already so I don't have to? I don't have a phone to run it on, anyway.

Edit: and I didn't even care for the original article whose headline boldly proclaimed that Flash was faster than HTML on mobile, not qualifying which mobile devices, nor the nature of the test.

[ Parent | on HN ]
Re:Not So Fast On That HTML5 Canvas vs. Flash Demo by dbrannan on Tuesday September 21, 2010 @04:30AM (#1711901)
From the many articles you see online, you'd think HTML5 was doing circles around Flash. It appears those statements are becoming less and less true, and even if this test isn't a 100% apples to apples comparison it is nice to see Flash performing well on a mobile.

Both of these technologies are in early stages of development (yes, Flash is mature on the desktop). I think in 1-2 years there will be less of a debate for which one is superior and just accept them for the content they provide.

[ Parent | on HN ]
Re:Not So Fast On That HTML5 Canvas vs. Flash Demo by buymorechuck on Tuesday September 21, 2010 @12:18PM (#1713129)
The drawing parts were close, but mobile profiles can greatly be affected by other parts (in this case the original FPS counter was consuming tons of CPU and battery). The optimized versions do a first approximation of what the Flash engine is doing under the hood, but are still fundamentally equivalent to the original drawing code.
[ Parent | on HN ]