I've been watching nim from the sidelines for a while now and am considering diving (again) in to test the waters especially considering the recent release announcement.
Just to add to the 'killer app' discussion above, I like shashlick's idea above. It is essentially what Julia is trying to do, but nim could be the compiled statically typed option along the same lines, but with a fast compiler, no JIT lag, and decent syntax. It could and should be billed as essentially solving the two language problem. I'd like to suggest that making nimscript easy to embed and possibly also offering some tooling support for workflows involving prototyping with nimscript and then later converting said code to regular nim. For game programming this could be very attractive for example. That and removing pain points from GC-less programming in nim, could make it both an attractive very high level and low level language. Just 2ยข from a friendly lurker.
