Add note about UTF-8 decoder
diff --git a/src/hb-buffer.cc b/src/hb-buffer.cc
index c20de77..a393b6b 100644
--- a/src/hb-buffer.cc
+++ b/src/hb-buffer.cc
@@ -607,6 +607,8 @@
   uint8_t c = *text;
   unsigned int mask, len;
 
+  /* TODO check for overlong sequences?  also: optimize? */
+
   UTF8_COMPUTE (c, mask, len);
   if (unlikely (!len || (unsigned int) (end - text) < len)) {
     *unicode = -1;