Updated the pull response to be more clear and short

docs/samples/response_pull.xml

@@ -1,18 +1,10 b''
 <?xml version="1.1" encoding="UTF-8" ?>
 <response>
     <status>success</status>
     <models>
-        <post>
-            <id>id1/1</id>
-        </post>
-        <post>
-            <id>id1/2</id>
-        </post>
-        <post>
-            <id>id2/1</id>
-        </post>
-        <post>
-            <id>id2/5</id>
-        </post>
+        <id>id1/1</id>
+        <id>id1/2</id>
+        <id>id2/1</id>
+        <id>id2/5</id>
     </models>
 </response>

docs/sync.markdown

@@ -1,29 +1,34 b''
 # INTRO #
 
 This document discribes the server sync protocol.
 
 # DRAFT PROTOCOL DESCRIPTION #
 
 The node requests other node's changes list since some time (since epoch if
 this is the start). The other node sends a list of post ids or posts in the
 XML or JSON format.
 
 Protocol version is the version of the sync api. Model version is the version
 of data models. If at least one of them is different, the sync cannot be
 performed.
 
 The node signs the data with its key. The receiving node saves the key at the
 first sync and checks it every time. If the key has changed, the info won't be
 saved from the node (or the node id must be changed).
 
 Each node holds a list of networks it wants to get info from.
 
 Each post has an ID in the unique format: node-id/post-id
 
 All requests pass a request type, protocol and model versions and a list of
 optional arguments used for filtering.
 
 # REQUESTS #
 Request types:
 * pull - pull the desired model ids
 * get - get models by ids
+
+# RESPONSES #
+* not supported - request is not supported
+* success - request was successfull
+* error - unexpected error