By Harry McCracken | Tuesday, May 11, 2010 at 9:34 pm
On Friday, Adobe evangelist Ryan Stewart demoed FlashPlayer 10.1 for Android at a Seattle conference. It crashed repeatedly and couldn’t deal with one site (Hulu) that an audience member asked to see. Given Flash’s already-shaky reputation at the moment, it wasn’t surprising that this led to some withering comments.
Stewart has published a sheepish blog post saying he was using an interim build of the software that wasn’t up to the stress test of a public demo. The post includes an embedded video of Stewart demoing several Flash sites on a Nexus One phone. And it…works. Quite well.
Here it is:
Me, I wasn’t that alarmed over Stewart’s Friday fiasco, nor am I all that swayed by his rebound. Having sat through thousands of tech demos over the past couple of decades, I’ve come to the conclusion that any correlation between the quality of a demo and the quality of a product is tenuous at best. I’ve seen products that knocked my socks off when demoed in a controlled environment, and which barely worked at all when they shipped. (A wise editor I once worked for told me to be especially skeptical about demos performed by the person who designed the product–in some cases that person is about the only one who can make a product sing.)
Bad demos are no more definitive than good ones: I’ve seen products that turned out to be entirely praiseworthy crash, spawn error messages, and otherwise misbehave. (There are, of course, two types of crashes: Those suffered by shipping products, and those that occur with beta products. The former are never a good sign; the latter are nothing to panic over. If a program didn’t crash, it probably wouldn’t be in beta.)
Here’s one piece of news I know is good: Adobe says it’s going to have a beta of FlashPlayer 10.0 for Android ready for Google’s I|O developer conference next week. Anyone who gets his or her hands on it will be in a far better position to have opinions about mobile Flash than anyone who hasn’t…