diff --git a/GSOC-Ideas.md b/GSOC-Ideas.md
index 56f234e..20e1942 100644
--- a/GSOC-Ideas.md
+++ b/GSOC-Ideas.md
@@ -16,4 +16,10 @@ Investigate speeding up core operations by vectorizing core functionality. This
 
 _Suggested by Eric Soroos, Mentor_
 
-Pillow decodes a lot of image formats, some of which are safer and better specified than others. Experience with [other projects](http://googleonlinesecurity.blogspot.com/2014/01/ffmpeg-and-thousand-fixes.html) has shown that there are many bugs lurking in otherwise functional code. Running a large corpus of valid, invalid, or otherwise fuzzed images while recording and investigating errors and crashes should improve the robustness of Pillow.  
\ No newline at end of file
+Pillow decodes a lot of image formats, some of which are safer and better specified than others. Experience with [other projects](http://googleonlinesecurity.blogspot.com/2014/01/ffmpeg-and-thousand-fixes.html) has shown that there are many bugs lurking in otherwise functional code. Running a large corpus of valid, invalid, or otherwise fuzzed images while recording and investigating errors and crashes should improve the robustness of Pillow.  
+
+### Coverage
+
+_Suggested by Jeff Breidenbach_
+
+Pillow decodes a lot of image formats, but there is room for more. In particular, JPEG 2000 is widely used in the motion picture industry and also by digital libraries. It would be handy to support this format, even if was on a read-only basis. Perhaps through a helper library such as libopenjpeg.
\ No newline at end of file