diff --git a/GSDoc/gsdoc.gsdoc b/GSDoc/gsdoc.gsdoc
index 9f044ae0e..ba10aaa37 100644
--- a/GSDoc/gsdoc.gsdoc
+++ b/GSDoc/gsdoc.gsdoc
@@ -108,7 +108,7 @@
 	(in the GNUstep project) need to devote a lot of time and effort
 	to maintaining documentation tools.
       </p>
-      <section>
+      <section id="gsdoc">
 	<heading>Overall document structure</heading>
 	<p>
 	  A GSDoc document consists of a <ref id="head">head</ref> and a
@@ -173,6 +173,33 @@
 	  </desc>
 	</deflist>
       </section>
+      <section id="head">
+	<heading>The document head</heading>
+	<p>
+	  The document head contains information about the document as
+	  a whole: its title, authors, version, modification date,
+	  copyright, and perhaps an abstract of its contents.
+	  The title and at least one author are the only parts of the
+	  document head that must be present.
+	</p>
+	<deflist>
+	<term><label id="title">title</label></term>
+	<desc>
+	  The title of the document.
+	</desc>
+	</deflist>
+      </section>
+      <section id="body">
+	<heading>The document body</heading>
+	<p>
+	  The document body contains the main part of the document, it
+	  consists of an optional front part (for contents pages, overview
+	  etc), a sequence of any number of chapters, and an optional back
+	  part (for indexes, appendices etc).  Normally, each of these three
+	  parts of the document woudl be expected to have their own separate
+	  page numbering schemes.
+	</p>
+      </section>
     </chapter>
 
     <chapter>
diff --git a/GSDoc/gsdoc.html b/GSDoc/gsdoc.html
index 774dac26c..2b0b2f617 100644
--- a/GSDoc/gsdoc.html
+++ b/GSDoc/gsdoc.html
@@ -23,9 +23,11 @@
 </ul>
 <li><a href="#cont-2">The gsdoc DTD and what it means</a>
 <ul>
-<li><a href="#cont-3">Overall document structure</a>
+<li><a href="#gsdoc">Overall document structure</a>
+<li><a href="#head">The document head</a>
+<li><a href="#body">The document body</a>
 </ul>
-<li><a href="#cont-4">The gsdoc conversion tool</a>
+<li><a href="#cont-3">The gsdoc conversion tool</a>
 </ul>
 <h2><a name="cont-0">Introduction</a></h2>
 <p>
@@ -113,11 +115,11 @@
 	(in the GNUstep project) need to devote a lot of time and effort
 	to maintaining documentation tools.
       </p>
-<h3><a name="cont-3">Overall document structure</a></h3>
+<h3><a name="gsdoc">Overall document structure</a></h3>
 <p>
 
-	  A GSDoc document consists of a <a href="#label-head">head</a> and a
-	  <a href="#label-body">body</a> wrapped inside the overall document
+	  A GSDoc document consists of a <a href="#head">head</a> and a
+	  <a href="#body">body</a> wrapped inside the overall document
 	  framework that looks like this -
 	</p>
 <a name="label-0">example</a>
@@ -148,7 +150,7 @@
  	</p>
 <dl>
 <dt>base
-<dt>
+<dd>
 	    This is optional, but recommended - it specifies the base name
 	    for the document.  When the document is tranlated to another
 	    format, the outut file name should be based on this - eg.
@@ -156,7 +158,7 @@
 	    output for this file would be <em>foo.html</em>.
 	  
 <dt>prev
-<dt>
+<dd>
 	    This optional attribute may be used as the name of a document
 	    that precedes this one in logical reading order.  If the
 	    converted output format of the document supports some sort of
@@ -164,7 +166,7 @@
 	    a link between the two documents.
 	  
 <dt>next
-<dt>
+<dd>
 	    This optional attribute may be used as the name of a document
 	    that follows this one in logical reading order.  If the
 	    converted output format of the document supports some sort of
@@ -172,7 +174,7 @@
 	    a link between the two documents.
 	  
 <dt>up
-<dt>
+<dd>
 	    This optional attribute may be used as the name of a document
 	    that is above this document in some sort of hierarchical
 	    structure (a contents list perhaps).
@@ -181,7 +183,32 @@
 	    a link between the two documents.
 	  
 </dl>
-<h2><a name="cont-4">The gsdoc conversion tool</a></h2>
+<h3><a name="head">The document head</a></h3>
+<p>
+
+	  The document head contains information about the document as
+	  a whole: its title, authors, version, modification date,
+	  copyright, and perhaps an abstract of its contents.
+	  The title and at least one author are the only parts of the
+	  document head that must be present.
+	</p>
+<dl>
+<dt><a name="title">title</a>
+<dd>
+	  The title of the document.
+	
+</dl>
+<h3><a name="body">The document body</a></h3>
+<p>
+
+	  The document body contains the main part of the document, it
+	  consists of an optional front part (for contents pages, overview
+	  etc), a sequence of any number of chapters, and an optional back
+	  part (for indexes, appendices etc).  Normally, each of these three
+	  parts of the document woudl be expected to have their own separate
+	  page numbering schemes.
+	</p>
+<h2><a name="cont-3">The gsdoc conversion tool</a></h2>
 <p>
 
 	The gsdoc tool is written in Objective-C and uses the GNUstep base
diff --git a/GSDoc/gsdoc.m b/GSDoc/gsdoc.m
index 438de9bfe..94ae0e6e5 100644
--- a/GSDoc/gsdoc.m
+++ b/GSDoc/gsdoc.m
@@ -1591,7 +1591,7 @@ loader(const char *url, const char* eid, xmlParserCtxtPtr *ctxt)
 	      NSLog(@"term without desc");
 	      return nil;
 	    }
-	  [text appendFormat: @"<dt>%@\r\n", [self parseDesc: node]];
+	  [text appendFormat: @"<dd>%@\r\n", [self parseDesc: node]];
 	  node = node->next;
 	}
       [text appendString: @"</dl>\r\n"];
@@ -1899,7 +1899,7 @@ loader(const char *url, const char* eid, xmlParserCtxtPtr *ctxt)
 	    else if (strcmp(node->name, "ref") == 0)
 	      {
 		NSString	*elem = [self parseText: node->childs];
-		NSString	*typ = [self getProp: "type" fromNode: node];
+//		NSString	*typ = [self getProp: "type" fromNode: node];
 //		NSString	*cls = [self getProp: "class" fromNode: node];
 		NSString	*ref = [self getProp: "id" fromNode: node];
 
@@ -1907,11 +1907,6 @@ loader(const char *url, const char* eid, xmlParserCtxtPtr *ctxt)
 		  {
 		    elem = ref;
 		  }
-		if (typ == nil)
-		  {
-		    typ = @"label";
-		  }
-		ref = [NSString stringWithFormat: @"%@-%@", typ, ref];
 		[text appendString: [self addLink: ref withText: elem]];
 	      }
 	    else if (strcmp(node->name, "uref") == 0)