diff --git a/README b/README
index 0f1253b..422afe1 100644
--- a/README
+++ b/README
@@ -13,7 +13,9 @@ make
 
 TODO
 
-* Proper handling of DQT tables that don't match the SSDV tables.
 * Allow the decoder to handle multiple images in the input stream.
-* Try adaptive huffman tables.
+* Experiment with adaptive or multiple huffman tables.
+* Support more than one subsampling configuration. (4 modes / 2 bits)
+* Quality setting (4 bit / 16 quality levels).
+* Add a callsign field to each packet.