Fixed an issue on the AES-CTR block mode

The previous version of the CTR encoding was incrementing the counter in a weird way, mixing up data from the previous block.
The current fix can correctly decrypt data encoded with AES-CTR using other libraries/languages (currently tested only with Python).

frontend/gamma/js/Clipperz/PM/DataModel/User.js

@@ -726,8 +726,8 @@ Clipperz.Base.extend(Clipperz.PM.DataModel.User, Object, {
 			header = {};
 			header['records']			= someHeaderPackedData['recordIndex']['records'];
 			header['directLogins']		= someHeaderPackedData['recordIndex']['directLogins'];
-			header['preferences']		= {'data': someHeaderPackedData['preferences']['data']};		//	this._serverData['header']['preferences'];		//	Clipperz.Base.evalJSON(this._serverData['header']['data'])['preferences'];		//	???????????
-			header['oneTimePasswords']	= {'data': someHeaderPackedData['oneTimePasswords']['data']};	//	this._serverData['header']['oneTimePasswords'];	//	Clipperz.Base.evalJSON(this._serverData['header']['data'])['oneTimePasswords'];	//	???????????
+			header['preferences']		= {'data': someHeaderPackedData['preferences']['data']};
+			header['oneTimePasswords']	= {'data': someHeaderPackedData['oneTimePasswords']['data']};
 			header['version']			= '0.1';
 
 			aResult['header']		= Clipperz.Base.serializeJSON(header);