diff --git a/lib/msg.py b/lib/msg.py
index 1f218db..7b0fa27 100644
--- a/lib/msg.py
+++ b/lib/msg.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python2
 # -*- coding: UTF-8 -*-
 # File: msg.py
-# Date: Sat Dec 20 18:38:59 2014 +0800
+# Date: Sun Dec 21 21:23:27 2014 +0800
 # Author: Yuxin Wu <ppwwyyxxc@gmail.com>
 
 from datetime import datetime
@@ -61,6 +61,7 @@ class WeChatMsg(object):
             url = soup.find('url').text
             if not url:
                 title = soup.find('title').text
+                # TODO sometimes fail to parse title
                 if not title:
                     print self.content
                     from IPython import embed; embed()
diff --git a/lib/render.py b/lib/render.py
index c5691b8..b40f48f 100644
--- a/lib/render.py
+++ b/lib/render.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python2
 # -*- coding: UTF-8 -*-
 # File: render.py
-# Date: Sat Dec 20 20:15:47 2014 +0800
+# Date: Sun Dec 21 21:21:53 2014 +0800
 # Author: Yuxin Wu <ppwwyyxxc@gmail.com>
 
 import os
@@ -90,7 +90,9 @@ class HTMLRender(object):
             bigimgpath = self.parser.imginfo.get(msg.msgSvrId)
             fnames = [k for k in [imgpath, bigimgpath] if k is not None]
             bigimg, smallimg = self.res.get_img(fnames)
-            assert smallimg
+            if not smallimg:
+                logger.warn("No image thumbnail found for {}".format(imgpath))
+                return fallback()
             # TODO do not show fancybox when no bigimg found
             return template.format(sender_label=sender,
                                    small_img=smallimg,
diff --git a/lib/res.py b/lib/res.py
index 80a97dd..6306a27 100644
--- a/lib/res.py
+++ b/lib/res.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python2
 # -*- coding: UTF-8 -*-
 # File: res.py
-# Date: Sat Dec 20 20:14:40 2014 +0800
+# Date: Sun Dec 21 21:22:44 2014 +0800
 # Author: Yuxin Wu <ppwwyyxxc@gmail.com>
 
 import glob
@@ -155,6 +155,7 @@ class Resource(object):
         if len(candidates) > 1:
             # annimation
             candidates = [k for k in candidates if not re.match('.*_[0-9]+$', k)]
+            # only one file is the gif in need, others are frames and cover
             assert len(candidates) == 1
         fname = candidates[0]
         return Resource.get_file_b64(fname), imghdr.what(fname)
diff --git a/lib/static/TP_EMOJI.html b/lib/static/TP_EMOJI.html
new file mode 100644
index 0000000..ad8d095
--- /dev/null
+++ b/lib/static/TP_EMOJI.html
@@ -0,0 +1,16 @@
+<div class="chatItem {sender_label}">
+  <div class="chatItemContent">
+    <img class="avatar" src="">
+    <div class="cloud cloudImg">
+      <div class="cloudPannel">
+        <div class="cloudBody">
+          <div class="cloudContent">
+            <span class="img_wrap">
+              <img class="zoomIn imageBorder" src="data:image/{emoji_format};base64,{emoji_img}"/>
+            </span>
+          </div>
+        </div>
+      </div>
+    </div>
+  </div>
+</div>