I agree. This is the real topic. Apologies for my IDE diversion.Nope. It s about real software engineering and the fact that JS is/was not designed for that.
And looking to the future ...
Look at the closure compiler which generates probably the most efficient JS out there. You wont get that level of efficiency without a static type system.
It s true that JS engine has gotten better and better(V8).
But the fastest code is the one that does not run right ?
Comparing browser JS speed to AS3 is really a straw man. Adobe is one company with many products, so the Flash runtime only gets a small portion of their resources. Compare that with the hundreds of millions of dollars that Mozilla, Google, and Microsoft have poured into maximizing the speed of JS.
So even if execution speed were the issue (which it isn't), it's not really a shock to find that JS may still be faster. The JS engines have had vastly more attention and optimization.
OO languages are an answer to this problem, encapsulation through classes and public/private access help programmers keep control as applications scale to 100,000 lines. This is the primary benefit of TypeScript.
Denial of this means you are claiming that software engineers of the last 30 years were engaged in a mass group delusion.
What I deny is that TypeScript, or solutions like it, solves that large a proportion of the engineering challenge as people in this thread claim. It is not a silver bullet, because there is no silver bullet. This will not solve run-time performance issues, because those are all DOM-based. This will not solve people's struggling to learn API's and functional domains. This will not solve issues with cross-browser layout. This will not solve browser-specific bugs. Finally, it will do very little to solve the biggest challenge of all in working with large teams and large codebases: communication between team members.
And now I've said all there is to say on this topic. A fine day to you all.
Scala & HTML5 & Ext fan