Flash, Silverlight just Stop-Gaps
atalasoft.com
atalasoft.com
The whole web space is just client/server distributed computing environment with relatively thin clients and high powered, cluster back ends. AS3 or JS is just as much a programming language as C or Java, particularly AS3. They're converging to the new JS standard anyway. I looked at the w3 HTML5 page and it still looks like HTML is the view component. My understanding of UI work is that it's general best to break it up into the view for displaying stuff, the model of what is in the view and the controller that mediates user action and model changes and keeps the view in synch. So that is one problem to address and if HTML5 is also trying to be a proper programming language, it shouldn't. Programming is about managing what data is in ram and how to manipulate that data, coordinate communications with other computers and signal changes to low level view components. Maybe AS3 and JS will merge, maybe HTML5 will give some scripting language as well but if it's not one of the EcmaScript varients, why are they reinventing that wheel? We need a solid view component and we need a good engine to render it in. The flash VM is a pretty good choice with a solid API, maybe it will standardize to something else in the future but it's what we have now. There's only so much that a typical view component can do, they're usually easy to swap out if your code is designed properly (or easy-ish typically). I think the flash VM is a stable, well supported environment and I'm quite skeptical that HTML5 will replace it anytime soon.
As an aside, I'm not sure what the article was talking about with those examples that are "not possible" in Flash, as I see nothing that couldn't be done. If it's just because they were transformed with CSS3, there's nothing stopping a CSS3 parser being written in AS3.
<em>"The point here is that we can't allow such a fundamental piece to the future of the Internet rely on plugins."</em>
He doesn't establish this conclusion. Recent Google examples (Gears, O3D) show that it's useful to be able to add features to browsers without requiring a complete browser change. Using common cross-browser functionality seems a benefit.
(I can't tell what "FUD" he sees in my blogpost, either. I spent time reading, but wish he had backed up his claims.)