Or rather, it WILL use them soon. Mac HW acceleration isn't fully deployed yet, is it?

It seemed to me like the article is, "Well yes there are these problem areas but there is an alpha that fixes them, so the entirely line of criticism is bunk." Obviously I do not agree.

It's good that Flash is finally getting comparable video playback performance. It's bad it took this long. And I also question why most people care. Flash, no flash... all I care about is video. If Flash is an impediment, let's dump it. Does that make a few Flash developers sad? TOO FUCKING BAD. This is an industry where change is a constant, if you wed your career to a technology you've made a fatal mistake.

"Or rather, it WILL use them soon. Mac HW acceleration isn't fully deployed yet, is it?"

Hence where available. And it wasn't available on the mac because Apple hadn't opened up the APIs to do it.

Your summary of the article is pretty poor -- that was one statement, the majority of the article was about the fact that critics often cite Flash as a CPU hog when comparing it to things that by default are not CPU hogs (static html).

Second, if video was the only thing Flash did, it would already be dumped. Flash does a lot more than that. The reason video is the linchpin is that it's taken innumerable years for standardized video access, whereas Flash provided it cross-platform. Sure you complain about Flash now, but the web wouldn't be where it is now without it.

"Hence where available. And it wasn't available on the mac because Apple hadn't opened up the APIs to do it."

VMWare and Parallels were able to use HW accelleration right? Why can't Adobe?

VMWare and Parallels do very different things than Flash does. My (uninformed) guess would be that it has to do with the security sandbox. The kinds of things you would run on VMWare (installable applications) is very different than what you typically run in Flash (whatever random dreck you happened to stumble into on the internet :P).

It's much easier to do hardware accelerated 3d drawing with a toolkit for an interpreter than it is to make a full virtualization layer properly do hardware acceleration.

Your opinion is indeed uninformed. Fortunately for you, correction is easy to get on the internet.

No. You misunderstood my verbiage. Rather, that Adobe has had more than decade to refine their algorithms and data structures so asking comparable performance from, ahem, far less than a decade's worth of engineering development is rather short-sided.

HTML5 etc. isn't starting from scratch -- the idea that Adobe has had extra years to refine their algorithms compared to potential competitors is pretty nonsensical,

Not for graphics rendering on the PC (and then only in a future release), which really is pathetic in 2010.

A) My statement was in reference to video playback.

B) Is there another cross-platform framework that does provide unilateral hardware acceleration for graphics rendering?

You're right about A, sorry.

In response to B, how about Adobe's own Shockwave? O3D is another. Java unfortunately requires external libraries (and thus there's a security prompt), but they could easily integrate them.

I think the lack of another cross platform framework speaks more to the futility of trying to build market share with a plug in rather than any technical reason.

Futility? The whole issue with Flash right now is that they do have market share, because they're the only one in the market. Either something has Flash (desktops, possibly Android), or it doesn't. And the non-Flash platforms are currently missing out on a significant portion of web content. So the idea that browser plugin as a 'futile' way of building market share seems like nonsense to me.

What I'm saying is that building a competing platform (with, for example, 3d acceleration) is futile not for technical reasons, but because you've got the mother of all chicken and egg problems trying to convince both end users and developers to use the platform.

O3D was already superior to Flash for making games, with I'm assuming less than 10 man years invested. Haxe is a superior compiler to Adobe's command line mxmlc, and it's written by one guy. If gaining traction wasn't so hard, I'm certain Flash would be in serious trouble (the Flash IDE being the only thing saving it).

OpenGL isn't really in the same class as Flash for this context -- we're not talking about Flash just as a method of rendering graphics, but as a content/application framework that can also render graphics. If OpenGL did the other stuff Flash does, we wouldn't be in this situation to begin with.