Changed the spec to say that array2d objects don't put padding at the

ends of the rows.

dlib/array2d/array2d_kernel_abstract.h

@@ -52,10 +52,9 @@ namespace dlib
                 of this object throw exceptions.
 
 
-                Finally, note that this object stores each row of data contiguously 
-                in memory, and the overall layout is in row major order.  However,
-                there might be padding at the end of each row.  To determine the
-                offset from one row to another you can use width_step(). 
+                Finally, note that this object stores its data contiguously and in 
+                row major order.  Moreover, there is no padding at the end of each row.
+                This means that its width_step() value is always equal to sizeof(type)*nc().  
         !*/
 
 
@@ -229,6 +228,12 @@ namespace dlib
                 - returns the size of one row of the image, in bytes.  
                   More precisely, return a number N such that:
                   (char*)&item[0][0] + N == (char*)&item[1][0].
+                - for dlib::array2d objects, the returned value
+                  is always equal to sizeof(type)*nc().  However,
+                  other objects which implement dlib::array2d style
+                  interfaces might have padding at the ends of their
+                  rows and therefore might return larger numbers.
+                  An example of such an object is the dlib::cv_image.
         !*/
 
     private: