WebAssembly has no access to any of the web API (DOM, ...) . WebAssembly cannot start without javascript as of right now. Flash died because of one platform not supporting it. The very same platform still doesn't support WebAssembly. As much as I love the hyper around WASM, I don't want to overhype it, otherwise, it will be a almost non-event as NPAPI.
I think that's an overly pessimistic view, which complements my overly optimistic view. :)
> WebAssembly has no access to any of the web API (DOM, ...)
WebAssembly can access those APIs via indirect, opaque shims today. Direct access is on the roadmap.
> WebAssembly cannot start without javascript as of right now.
We're still hammering out semantics on the HTML side, but WebAssembly modules can define a `start` entrypoint that is automatically invoked when the module is loaded. Combine that with ES6's `<script type="module">` and you've got WASM that can load and begin executing without JS.
> Flash died because of one platform not supporting it. The very same platform still doesn't support WebAssembly.
Apple is a party to WebAssembly's development and a co-signatory on the linked announcement that the WebAssembly binary format and Web APIs are complete. This doesn't mean they'll necessarily ship in a timely manner, or at all, but it's a damn good sign.
> I don't want to overhype it
That's fair. The world won't change overnight, but it's clear that the world is changing in fundamental ways. The initial changes may be insignificant, but they'll compound over time and as WebAssembly matures.
JavaScript isn't dead, and WebAssembly won't kill it, but it will dramatically change the landscape. In a similar vein, C will never kill Python, but I'd be shocked if most moderately complex Python applications didn't depend on a module implemented in C somewhere in their stack.
