diff --git a/asklyphe-frontend/src/spellcheck.rs b/asklyphe-frontend/src/spellcheck.rs
index 5f481c3..732c6be 100644
--- a/asklyphe-frontend/src/spellcheck.rs
+++ b/asklyphe-frontend/src/spellcheck.rs
@@ -6,6 +6,7 @@ use std::sync::Mutex;
 
 // how to generate words.txt:
 // clone https://github.com/en-wl/wordlist && cd wordlist
+// make
 // ./scowl wl --deaccent > words0.txt
 // filtered with this python script:
 // -----------------------------------
@@ -195,8 +196,8 @@ fn levenshtein_distance(a: &str, other: &str) -> usize {
 				// TODO: make addition/subtraction 1 more expensive than replacement, presumably by adding '+ 1' to 2/3 of these
 				// motivation: honex from bee movie script is turned into hone instead of honey, this will also generally improve results & is what wikipedia says to do (best reason)
 				dist[j] = 1 + cmp::min(
-					dist.get(j - 1).unwrap(),
-					cmp::min(dist_prev.get(j).unwrap(), dist_prev.get(j - 1).unwrap()));
+					*dist.get(j - 1).unwrap() + 1,
+					cmp::min(*dist_prev.get(j).unwrap() + 1, *dist_prev.get(j - 1).unwrap()));
 			}
 		}
 		mem::swap(&mut dist, &mut dist_prev);