4 ms·
Your assumption is wrong. Lots of really useful math has 0 or 1 implementations. A Julia one may well be first. (Saying this as someone who's done several first
by gspr 5y ago
Your assumption is wrong. Lots of really useful math has 0 or 1 implementations. A Julia one may well be first. (Saying this as someone who's done several first implementations, but haven't coded a single line of Julia)