On 7/20/26 18:40, John Cowan wrote:
> This is another silent breaking change to historic (sin 0) => 0.0 etc.,
> but hopefully less problematic, since user-level tests for exactness are
> rare.

"Historic"?  I can imagine that some Schemes do/have done this, because
it's what C and other languages do automatically, but I don't see that
this gives it any intrinsic superiority to (sin 0) => 0.