Hey, Karl has asked me to join in the discussion because I am working on some software that uses Docbook2X to convert between DocBook and PDF/PS/DVI etc.
I can say that in my experience, Docbook2X is the most mature/complete DocBook-Texinfo free software tool available. That it involves a transformation into Texinfo is a huge bonus because instead of trying (and failing) to master such formats as PDF/PS/DVI it concentrates on getting Texinfo produced reliably and lets you hook into the already incredibly mature Texinfo toolchain. I would like to note that I exlucde any Java tools because the freeness of the required Java runtime has always been an issue. Anyway, onwards to your issues so far: 1) Every time you get a "template not matched" it is because you are using a feature that Docbook2x does not support, which usualy means something version 5 specific. You can ignore these if you like, or transform using XSLT, it's usually not fatal. You should take a look at the tool xsltproc for this. 2) From the error messages that Docbook2x seemed to be giving you it sounds like you have a set of files (one for each chapter, say) using xincludes and XML entity references. I'm actually playing devils advocate, I've used the Subversion book for testing my own application, so I know this is the case. Heh. Before you run the book through Docbook2x you need to canonicalise the XML, replace any XML entity references and resolve the xincludes. This should give you one monolithic XML file that you can feed into Docbook2X without any problem. I suggest you take a look at the tool xmllint for this. A sample command: xmllint --c14n --noent --nocdata --xinclude book.xml > monolithic-book.xml HTH, -- Noah Slater <http://bytesexual.org/> "Creativity can be a social contribution, but only in so far as society is free to use the results." - R. Stallman