diff --git a/addressbook/inc/class.addressbook_merge.inc.php b/addressbook/inc/class.addressbook_merge.inc.php
index 8d9e761600..cdfe380db6 100644
--- a/addressbook/inc/class.addressbook_merge.inc.php
+++ b/addressbook/inc/class.addressbook_merge.inc.php
@@ -46,6 +46,12 @@ class addressbook_merge	// extends bo_merge
 	 */
 	static function is_implemented($mimetype,$extension=null)
 	{
+		static $zip_available;
+		if (is_null($zip_available))
+		{
+			$zip_available = check_load_extension('zip') &&
+				class_exists('ZipArchive');	// some PHP has zip extension, but no ZipArchive (eg. RHEL5!)
+		}
 		switch ($mimetype)
 		{
 			case 'application/msword':
@@ -54,11 +60,11 @@ class addressbook_merge	// extends bo_merge
 			case 'text/rtf':
 				return true;	// rtf files
 			case 'application/vnd.oasis.opendocument.text':
-				if (!check_load_extension('zip')) break;
+				if (!$zip_available) break;
 				return true;	// open office write xml files
 			case 'application/vnd.openxmlformats-officedocument.wordprocessingml.document':
 			case 'application/vnd.openxmlformats-officedocument.wordprocessingml.d':	// mimetypes in vfs are limited to 64 chars
-				if (!check_load_extension('zip')) break;
+				if (!$zip_available) break;
 				return true;	// ms word xml format
 			default:
 				if (substr($mimetype,0,5) == 'text/')
@@ -269,8 +275,9 @@ class addressbook_merge	// extends bo_merge
 			return false;
 		}
 		list($contentstart,$contentrepeat,$contentend) = preg_split('/\$\$pagerepeat\$\$/',$content,-1, PREG_SPLIT_NO_EMPTY);  //get differt parts of document, seperatet by Pagerepeat
-		if ($mimetype == 'application/vnd.oasis.opendocument.text' && count($ids) > 1) {
-			//for odt files we have to slpitt the content and add a style for page break to  the style area
+		if ($mimetype == 'application/vnd.oasis.opendocument.text' && count($ids) > 1)
+		{
+			//for odt files we have to slpit the content and add a style for page break to  the style area
 			list($contentstart,$contentrepeat,$contentend) = preg_split('/office:body>/',$content,-1, PREG_SPLIT_NO_EMPTY);  //get differt parts of document, seperatet by Pagerepeat
 			$contentstart = substr($contentstart,0,strlen($contentstart)-1);  //remove "<"
 			$contentrepeat = substr($contentrepeat,0,strlen($contentrepeat)-2);  //remove "</";
@@ -418,8 +425,9 @@ class addressbook_merge	// extends bo_merge
 				copy($content_url,$archive);
 				$content_url = 'zip://'.$archive.'#'.($content_file = 'content.xml');
 				break;
-			case 'application/vnd.openxmlformats-officedocument.wordprocessingml.document':
 			case 'application/vnd.openxmlformats-officedocument.wordprocessingml.d':	// mimetypes in vfs are limited to 64 chars
+				$mime_type = 'application/vnd.openxmlformats-officedocument.wordprocessingml.document';
+			case 'application/vnd.openxmlformats-officedocument.wordprocessingml.document':
 				$archive = tempnam($GLOBALS['egw_info']['server']['temp_dir'], basename($document,'.dotx').'-').'.dotx';
 				copy($content_url,$archive);
 				$content_url = 'zip://'.$archive.'#'.($content_file = 'word/document.xml');
@@ -437,7 +445,7 @@ class addressbook_merge	// extends bo_merge
 			if ($zip->close() !== true) throw new Exception("!ZipArchive::close()");
 			unset($zip);
 			unset($merged);
-			if (file_exists('/usr/bin/zip'))	// fix broken zip archives generated by current php
+			if (file_exists('/usr/bin/zip') && version_compare(PHP_VERSION,'5.3.1','<'))	// fix broken zip archives generated by current php
 			{
 				exec('/usr/bin/zip -F '.escapeshellarg($archive));
 			}