Spec: clarification needed for remainder operator on float operands

[peterhuene]

Currently the remainder % operator (i.e. modulo) is accepting of Int and Float as operand types.

For two Int types, there is no question as to how the operation should be performed as it involves only integer division, e.g. a - ((a / b) * b), where a / b naturally truncates.

However, for Float operands or Int operands implicitly coerced to Float because the other operand is Float, floating point division is involved and there are several different methods for performing the operation.

I would assume the intention is to define the operator in terms of truncation to align with most programming languages, but some notable languages are flooring instead (e.g. Python, Ruby), so an explicit clarification is probably required for the WDL specification.