Fix spelling mistakes and revision block formatting

* superseeding => superseding
* meta data => metadata
* seperate => separate

xep-0398.xml

@@ -32,9 +32,7 @@
     <version>0.2.1</version>
     <date>2018-08-27</date>
     <initials>egp</initials>
-    <remark><ul>
-      <li>Fix typo.</li>
-    </ul></remark>
+    <remark>Add missing article "the".</remark>
   </revision>
   <revision>
     <version>0.2.0</version>
@@ -46,11 +44,11 @@
     <version>0.1.0</version>
     <date>2017-12-18</date>
     <initials>dg</initials>
-    <remark><p>First draft.</p></remark>
+    <remark>First draft.</remark>
   </revision>
 </header>
 <section1 topic='Introduction' anchor='intro'>
-  <p>&xep0084; and &xep0153; are usually considered to stand in competition with each other. <cite>XEP-0084</cite> even talks about superseeding <cite>XEP-0153</cite> in the future. While <cite>XEP-0084</cite> provides a more efficient interface to upload avatars by seperating meta data and data (thus saving the client from having to download its own avatar on every connect) it has the significant downside of not working with &xep0045;.</p>
+  <p>&xep0084; and &xep0153; are usually considered to stand in competition with each other. <cite>XEP-0084</cite> even talks about superseding <cite>XEP-0153</cite> in the future. While <cite>XEP-0084</cite> provides a more efficient interface to upload avatars by separating metadata and data (thus saving the client from having to download its own avatar on every connect) it has the significant downside of not working with &xep0045;.</p>
   <p>Server implementations can aid to resolve this conflict by automatically putting avatars uploaded with <cite>XEP-0084</cite> into <cite>XEP-0153</cite> storage and vice versa. This allows clients to use the more efficient <cite>XEP-0084</cite> for uploading avatars and <cite>XEP-0153</cite> to retrieve avatars in Multi-User Chats.</p>
 </section1>
 <section1 topic='Discovery' anchor='disco'>