[RFC] Adding better support for higher precision floating-point

I knew this (using builtin types) was being discussed, but wasn’t sure that a decision had been made. Thanks for confirming this direction.

Sounds like adding f80 and f128 is the preferred approach. Good. :slight_smile: