The web runtime should try to cache Runes where possible

[Michael-F-Bryan]

As a convenience, the web runtime's builder() can accept the URL of a Rune. This is awesome because you don't have to call fetch() yourself, but can also be a massive pain when you need to download the same 50MB Rune from the internet every time you reload.

Inside Builder.build() we should use the browser's caching functionality to avoid re-downloading the same file every time.

@Ge-te and @binaryanthems, do you know what the best way to do this is?