I'm trying to get xml.sax to interpret a file that begins with… <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http:// www.w3.org/TR/html4/loose.dtd"> After a while I get... http://www.w3.org/TR/html4/loose.dtd:31:2: error in processing external entity reference …although… time curl http://www.w3.org/TR/html4/loose.dtd …gives… real 0m26.888s user 0m0.006s sys 0m0.013s Is this a rookie mistake? Should I expect a python SAX parser to incorporate entities from a remote DTD into it's parsing interpretation?