This might be useful for some document processing I need to do. Tried it out on a simple rtf file. The docbook output left rtf tags in the output (it says experimental, so I guess that's OK). The standard XML output seemed OK, but seemed to have less structure than could be obtained from the document. The problem was that the RTF file was left open And a huge chunk of my memory is gone after the program has quit. I guess I'll be putting this one in the trash...