diff --git a/CHANGELOG.md b/CHANGELOG.md
index ecd5552..14c8f07 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -55,9 +55,9 @@
   when the label runs over 999 characters, where it read them as a link or a definition. CommonMark
   caps a label at 999 characters, an escape counting as two.
 - Every reader leaves a link or a definition literal when its unbracketed destination holds a `(`
-  that is neither escaped nor closed by a `)` before the space ending the destination, where it
-  read it as a link: `[a](b(c "t")` is now text, as in CommonMark. Markdown the writers wrote reads
-  as before, since they escape such a parenthesis.
+  that is neither escaped nor closed, as CommonMark requires: `[a](b(c "t")` is now text, where it
+  read as a link to `b(c`. Markdown the writers wrote reads as before, since they escape such a
+  parenthesis.
 - The README compares the package with seven other ADF converters, each checked against its
   published version.
 
diff --git a/src/markdown/commonmark/link-syntax.test.ts b/src/markdown/commonmark/link-syntax.test.ts
index 10fc996..7455326 100644
--- a/src/markdown/commonmark/link-syntax.test.ts
+++ b/src/markdown/commonmark/link-syntax.test.ts
@@ -28,7 +28,7 @@ test('an unbracketed destination ends at its first unmatched `)`', () => {
   assert.deepEqual(targets('[a](b\\)c)'), [{ destination: 'b)c', length: 6 }])
 })
 
-test('a destination holding an unbalanced unescaped `(` at a space or the end reads as none', () => {
+test('an unbracketed destination with an unescaped `(` still open where it stops reads as none', () => {
   assert.deepEqual(targets('[a](b(c "t")'), [undefined])
   assert.deepEqual(targets('[a](b(c "t") [d](e)'), [undefined, { destination: 'e', length: 3 }])
   assert.deepEqual(targets('[a](b(c) "t")'), [{ destination: 'b(c)', length: 10, title: 't' }])
diff --git a/src/markdown/commonmark/link-syntax.ts b/src/markdown/commonmark/link-syntax.ts
index bcaba70..627d8a9 100644
--- a/src/markdown/commonmark/link-syntax.ts
+++ b/src/markdown/commonmark/link-syntax.ts
@@ -72,7 +72,7 @@ export function targetMemo(text: string): TargetMemo {
   return { destinationEnds: undefined, tails: new Map(), text }
 }
 
-// `ends[start]` is where an unbracketed destination from `start` stops, filled in one pass for every start a target can have: just after a `(`, a space or a control character, and 0, where `readDestination`'s run begins. At a stop only the innermost open start has its parentheses balanced, so the starts below it keep no end.
+// `ends[start]` is where an unbracketed destination from `start` stops, for every start a target can have: just after a `(`, a space or a control character, and 0, where `readDestination`'s run begins. A destination stops at a space, a control character or the end of the text. Of the starts on the `open` stack only the innermost has its parentheses balanced there, so every start below it keeps 0, which reads as no destination.
 function destinationEnds(text: string): Int32Array {
   const ends = new Int32Array(text.length + 1)
   let open = [0]