README: update go version requirement and README

Author: ALTree

README.md

@@ -1,16 +1,9 @@
-### Floats 
-
-Package bigfloat provides the implementation of a few additional operations (square root, exponentiation, natural logarithm, exponential function) for the standard library `big.Float` type.
+Package bigfloat provides arbitrary-precision natural logarithm and
+exponentiation for the standard library's `big.Float` type.
 
 [![GoDoc](https://godoc.org/github.com/ALTree/bigfloat?status.png)](https://godoc.org/github.com/ALTree/bigfloat)
 
-#### Install
-
-```
-go get github.com/ALTree/bigfloat
-```
-
-Please note that `bigfloat` requires Go >= 1.5 (since the `big.Float` type is not available in previous versions). 
+The package requires Go 1.10 or newer.
 
 #### Example
 
@@ -24,31 +17,22 @@ import (
 	"github.com/ALTree/bigfloat"
 )
 
-// In this example, we'll compute the value of the
-// transcendental number 2 ** √2, also known as
-// the Gelfond–Schneider constant, to 1000 bits.
+// We'll compute the value of the transcendental number 2^√2, also
+// known as the Gelfond–Schneider constant, to 1000 bits.
 func main() {
-	// Work with 1000 binary digits of precision.
-	const prec = 1000
-
+	const prec = 1000 // in binary digits
 	two := big.NewFloat(2).SetPrec(prec)
+	sqrtTwo := new(big.Float).SetPrec(prec).Sqrt(two)
 
-	// Compute √2.
-	// Sqrt uses the argument's precision.
-	sqrtTwo := bigfloat.Sqrt(two)
-
-	// Compute 2 ** √2
 	// Pow uses the first argument's precision.
-	gsc := bigfloat.Pow(two, sqrtTwo)
-
-	// Print gsc, truncated to 60 decimal digits.
-	fmt.Printf("gsc = %.60f...\n", gsc)
+	gsc := bigfloat.Pow(two, sqrtTwo) // 2^√2
+	fmt.Printf("gsc = %.60f\n", gsc)
 }
 ```
 
-outputs
+outputs:
 ```
-gsc = 2.665144142690225188650297249873139848274211313714659492835980...
+gsc = 2.665144142690225188650297249873139848274211313714659492835980
 ```
 
 #### Documentation