[Pythoncard-users] SourceForgeTracker now uses DOM parsing

The big problem all along with the SourceForgeTracker sample appears to be
that SF is not generating valid XML documents. Anyway, I updated the code
this weekend and unless Mark or someone with more XML/Unicode knowledge than
me (aka everyone) suggests further changes I'm done with it for now.

Thanks to Mark Pilgrim for the DOM parsing code.
http://diveintopython.org/kgp_divein.html
Thanks to Mark Pilgrim and Martin Martin v. Loewis for help
with XML and UNICODE issues.
http://aspn.activestate.com/ASPN/Mail/Message/xml-sig/967244

It still doesn't download the Python Bug Tracker items because of an open
bug in the XML generator SF is using; the server runs out of memory and
chokes before the XML can be sent.

ka