Fallbacks and future API

[jkrems]

Even though import maps may drop fallbacks in the initial version, I still believe that being able to feature detect and/or load fallbacks is important. Adding exports as mentioned in other issues (#3) makes this harder. Could the proposal call out alternatives for future API extensions? My current assumption is that import 'lib:uuid/UUID' would be easier to handle than a named export since simple string-based fallbacks could catch it already.

[littledan]

I think there's a lot of possible design space for adding features to import maps. I don't think we're at the point where we should be designing libraries to cope with their restrictions.