diff --git a/src/main.rs b/src/main.rs
index 5d2d165..b013654 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -1,5 +1,78 @@
+use std::error::Error;
 use std::env;
 
-fn main() {
-	let args = env::args().collect::<Vec<_>>();
+fn main() -> Result<(), Box<dyn Error>> {
+	let args = env::args().skip(1).collect::<Vec<_>>();
+	for arg in args {
+		let binary_value = parse(&arg.replace(' ', ""))?;
+		println!("binary: {:032b}", binary_value);
+		let transmuted = f32::from_bits(binary_value); // clippy was smart enough to notice the transmute to a f32
+		println!("transmuted: {:?}", transmuted);
+		println!("interpreted: {:?}", interpret_single(binary_value));
+	}
+	Ok(())
+}
+
+fn parse(binary_number: &str) -> Result<u32, Box<dyn Error>> {
+	if binary_number.len() != 32 {
+		return Err("argument not 32 bits".into());
+	}
+	let mut value: u32 = 0;
+	for (idx, digit) in binary_number.chars().enumerate() {
+		match digit {
+			'0' => {},
+			'1' => value += 1 << (31-idx),
+			x => return Err(format!("argument contains more than 0s and 1s: {:?}", x).into())
+		}
+	}
+	Ok(value)
+}
+
+#[test]
+fn parse_test() {
+	assert_eq!(parse("10110000111111110101010110101010").unwrap(), 0b10110000111111110101010110101010);
+}
+
+fn interpret_single(float: u32) -> f32 {
+	let sign = float.rotate_left(1) & 1;
+	let sign = if sign == 1 { -1.0 } else { 0.0 };
+	let characteristic = float.rotate_left(9) & 0b1111_1111;
+	let mantissa = float & 0b1111_1111_1111_1111_1111_111;
+
+	if characteristic == 0 {
+		if mantissa != 0 {
+			eprintln!("can't process denormalized floats (yet)");
+			return 0.0;
+		} else {
+			return sign * 0.0;
+		}
+	} else if characteristic == 255 {
+		if mantissa != 0 {
+			return f32::NAN;
+		} else {
+			return sign * f32::INFINITY;
+		}
+	}
+
+	let mut power = 2.0f32.powi(characteristic as i32 - 127);
+	let mut value = sign * power;
+	for idx in 0..23 {
+		let digit = (mantissa >> (22 - idx)) & 1;
+		power /= 2.0;
+		if digit == 1 {
+			value += sign * power;
+		}
+	}
+	value
+}
+
+#[test]
+fn interpret_single_test() {
+	let tests = [
+		(0b1100_0000_1110_1000_0000_0000_0000_0000, -7.25),
+		(0b1011_1111_0101_0000_0000_0000_0000_0000, -0.8125)
+	];
+	for test in &tests {
+		assert_eq!(interpret_single(test.0), test.1);
+	}
 }