For one, WebAssembly is designed as a target for many languages. An LLVM backend is in the works, which means being able to write front-end code in one of proba

For one, WebAssembly is designed as a target for many languages. An LLVM backend is in the works, which means being able to write front-end code in one of probably 20-30 languages. I'm personally really hoping we get to the point where I can write my front-end in Rust. Second, Java GUI toolkits all sucked and produced ugly UIs. I know some developers liked Swing and people claim it can be made to look native, but I've used hundreds of Swing apps and they never lived up to that promise. For better or for worse, HTML+CSS works well enough and people understand it, so WebAssembly's ability to interact with the DOM is a massive improvement over what was available to Applets. Third, Sun's ability to secure the JVM plugin was, to be kind, pathetic. WebAssembly should be better since it should inherit a lot of what we already have for JavaScript. For better or for worse, the JVM was just never designed to protect users from the people writing Java code whereas all JavaScript engines were built from the ground up with that in mind. WebAssembly will be built with that same mindset. Fourth, an open specification vs a closed spec controlled by a single company. There's probably more reasons, but that covers most of what I hated about Applets that I'm hoping won't be the case with WebAssembly.