Parser crashes when using non-ASCII characters in IDs

As I'm from Germany, I wanted to use IDs containing
german umlauts (e.g. ... ID="Einführung").
Shouldn't this work if I use a encoding attribute for the XML file like
'<?xml version="1.0" encoding="ISO-8859-1" ?>' ?
 
Currently the parser crashes with the following error:
 


Fatal Error Messages

FatalError: ARP internal error[Line = -1, Column = -1]


Warning Messages

Warning: {W111} Failed to parse XML.[Line = -1, Column = -1]


Parser Loading Error

Exception parsing: http://www.kbs.uni-hannover.de/Lehre/KI1/ki1.rdf:
com.hp.hpl.jena.rdf.arp.CoRoutineDeathException 
 
 
 
On request I'm glad to provide a small sample RDF file.
 
regards,
Wolf Siberski