Trying to find out what makes the markdown go bizarre on linebreaks....

universalfilter/tablemanager/implementation/README.md

@@ -66,7 +66,8 @@ This transformation will be explained by a series of examples. First lets assume
 go deeper into an object. (e.g. we do not zoom in on a piece of the object we want to query.) The 
 return value I get from a the ResourceModel will either be an stdClass (object) or an array.
 
-Case 1) We get an object.
+Case 1) We get an object.  
+
     In case of an object, the names of the datamembers are important, and have to be used, not thrown away. Ofcourse
     the values of these datamembers are important, and have to be saved as well. This will result in a somewhat hash-map
     like datastructure ( e.g. { field1:"value1", field2:"value2" } ).  
@@ -151,9 +152,10 @@ Case 2) We got an array.
         </tr>
        </table>              
 
-    With associative arrays the indexes, or more correctly keys, are probably also important. Note that we don't save the information of "keys" in transforming
-    numerical arrays to tables. To save the "key" information on which a certain value is mapped, we add an extra "index" column to the resulting table. The rest
-    of the info is transformed to a table in the exact same way as numerical arrays are being transformed.
+    With associative arrays the indexes, or more correctly keys, are probably also important.  
+    Note that we don't save the information of "keys" in transforming numerical arrays to tables.  
+    To save the "key" information on which a certain value is mapped, we add an extra "index" column to the resulting table.  
+    The rest of the info is transformed to a table in the exact same way as numerical arrays are being transformed.  
 
        e.g.: 
            [