@@ -102,15 +102,14 @@ extension Augmented {
102
102
/// error from that computation rounded to the closest representable
103
103
/// value.
104
104
///
105
- /// Unlike `Augmented.sum(large:, small:)`, the magnitude of the summands
106
- /// does not matter. `a.magnitude` might as well be smaller than
107
- /// `b.magnitude` – and vice versa. However, it is recommended to only use
108
- /// this function over `Augmented.sum(large:, small:)` in cases where the
109
- /// ordering of the summands magnitude is unknown at compile time. In cases
110
- /// where either of the summands magnitude is known to be greater than or
111
- /// equal the magnitude of the other summand, use
112
- /// `Augmented.sum(large:, small:)` over this function; as it faster to
113
- /// calculate.
105
+ /// Unlike `Augmented.sum(large: a, small: b)`, the magnitude of the summands
106
+ /// does not matter and `a.magnitude` might as well be strictly less than
107
+ /// `b.magnitude`. However, it is recommended to only use this function over
108
+ /// `Augmented.sum(large: a, small: b)` in cases where the ordering of the
109
+ /// summands magnitude is unknown at compile time. In cases where either of
110
+ /// the summands magnitude is guaranteed to be greater than or equal the
111
+ /// magnitude of the other summand, use `Augmented.sum(large: a, small: b)`
112
+ /// over this function; as it faster to calculate.
114
113
///
115
114
/// Unlike `Augmented.product(a, b)`, the rounding error of a sum can
116
115
/// never underflow. However, it may not be exactly representable when
0 commit comments