Ultimately, a VM is just another language with lousy readability.
I agree its a very hard problem. If I was starting from scratch I would probably look a lot more on vector operations than someone would have 10 years ago. Plus the manner of parallelism a person choses would seriously influence language implementors.
Calling JavaScript "flexible bytecode" is like saying that all Turing-complete programming languages are the same, or that all objects capable of driving a nail into a piece of wood are equivalent. It's true in a very facile way. JavaScript can be treated kind of like bytecode, and all Turing-complete languages can be treated as a simulation of a Turing machine, but the details that these rough descriptions miss are more numerous and more significant than the details that actually went into forming the generalization.
"The LLVM code representation is designed to be used in three different forms: as an in-memory compiler IR, as an on-disk bitcode representation (suitable for fast loading by a Just-In-Time compiler), and as a human readable assembly language representation. This allows LLVM to provide a powerful intermediate representation for efficient compiler transformations and analysis, while providing a natural means to debug and visualize the transformations. The three different forms of LLVM are all equivalent. This document describes the human readable representation and notation." http://llvm.org/docs/LangRef.html#abstract
Why, and who would you pick?
1. User opens a page with scripts in LanguageX. The page links to some kind of VM configuration file. 2. Browser loads that file, which enables it to parse LanguageX. Browser caches that file. 3. From now on, the browser can execute scripts in LanguageX.
Although, this particular idea might be the case when too much flexibility would lead to issues in the long run.
I realize this would be very difficult and introduce horrible compatibility issues for a while, but sometimes I think, "what if we could start over?" What if Chrome and Firefox started supporting Python or Ruby or a custom, well-designed language for web scripting? Could it gain enough momentum to replace Javascript?
Actually, the easiest candidate would be for browsers to start natively supporting Coffeescript. But imagine a "dream team" designing a language just for the web, from scratch: Douglas Crockford, John Resig, Guido van Rossum, Yukihiro Matsumoto...
Is this a "it's crazy but it just might work" scenario, or is it just crazy?