Having retrieved the current code from the CVS I see no difference from my old version. No matter, it was a useful exercise. I realize there are problems with the current code: - Special characters (&,<,>) are not escaped by the toxml() method - there is no way to specify encoding for toxml or for display purpose in general The first is a trivial issue, but I am not sure how (or where) to solve the second one. It gets pretty annoying if at the management interface one types: <?xml version="1.0" encoding="iso-8859-1"?> <root> öäå </root> but after clicking on the Change button gets: <?xml version="1.0"?> <root> öÀå </root> As I understand it, the result is UTF-16. In most cases though, I would like to output it in 'iso-8859-1'. I think that XMLDocument should remember the encoding used to create the DOM and use that encoding to produce xml. Perhaps toXML() can take an encoding parameter and there is a getEncoding() method, too. So one can write: mydoc.toXML(encoding=mydoc.getEncoding()) I can imagine making this the default, as well. What do you guys think? And what is the policy on bringing new library dependencies? I am thinking about using more of xml-sig's stuff. Regards, Gabor