Solve both parts of day 11 in a single pass.

Weirdly, this has substantially increased the average runtime. I've occasionally gotten down to the 18 ms mark with this
build, but more often I get around 36. Not entirely sure what's going on there, but I have to assume that more runs of
the previous iteration would also have shown this higher average, and I just got lucky the first time.

Still, this does now take a single pass over the input, and the best recorded time was precisely equal to the only
recorded time of the two-pass solution, so I have to assume that the bulk of the runtime had to do not with walking the
hex positions but with parsing the input. Fair enough.

$ cargo check && cargo build --release && time target/release/day11 && wc -c input.txt
   Compiling day11 v0.1.0 (file:///mnt/d/Users/coriolinus/Documents/Projects/adventofcode.com/adventofcode-2017/day11)
    Finished dev [unoptimized + debuginfo] target(s) in 0.58 secs
    Finished release [optimized] target(s) in 0.0 secs
Dist to origin: 720
Max dist to origin: 1485

real    0m0.037s
user    0m0.000s
sys     0m0.016s
21533 input.txt

Author: coriolinus

day11/src/main.rs

@@ -13,15 +13,13 @@ fn main() {
             .collect()
     }).expect("Problem parsing input")
     {
-        let position: HexPosition = directions.iter().sum();
-        println!("Dist to origin: {}", position.min_steps_to_origin());
-
         let mut position = HexPosition::new();
         let mut max_dist = 0;
         for direction in directions.iter() {
             position = position + direction;
             max_dist = max(max_dist, position.min_steps_to_origin());
         }
+        println!("Dist to origin: {}", position.min_steps_to_origin());
         println!("Max dist to origin: {}", max_dist);
     }
 }