I tried to send this text to the
+
I am quite new to docbook processing, and am climbing a steep +learning curve at the moment.
+ +To give you some background, I am working on a Norwegian +translation of the book Free Culture by Lawrence Lessig, and I use +docbook to handle the process. The files to build the book are +available from +github. +The book got around 400 pages with parts, images, footnotes, tables, +index entries etc, which has proven to be a challenge for the free +software docbook processors. My build platform is Debian GNU/Linux +Squeeze.
+ +I want to build PDF, EPUB and HTML version of the book, and have +tried different tool chains to do the conversion from docbook to these +formats. I am currently focusing on the PDF version, and have a few +problems.
+ +-
+
+
- Using dblatex, the <part> handling is not the way I want to, + as </part> do not really end the <part>. (See + BTS report #683166), the + xetex backend (needed to process UTF-8) give incorrect hyphens in + index references spanning several pages (See + BTS report #682901), and + I am unable to get the norwegian template texts (See + BTS report #682936). + +
- Using straight xmlto fail with some latex error (See + BTS report + #683163). + +
- Using xmlto with the fop backend fail to handle images (do not + show up in the PDF), fail to handle a long footnote (overlap + footnote and text body, see + BTS report #683197), and + fail to create a correct index (some lack page ref, and the page + refs listed are not right). + +
- Using xmlto with the dblatex backend behave like dblatex. + +
- Using docbook-xls with xsltproc + fop have the same footnote and + index problems the xmlto + fop processing. + +
So I wonder, what would be the best way to create the PDF version +of this book? Are some of the bugs found above solved in new or +experimental versions of some docbook tool chain?
+ +What about HTML and EPUB versions?
+ +