Generating PDF: fop, passivetex and xfc (fo2rtf) reports

[Jean Jordaan <jean@upfrontsystems.co.za>]

Hi all

An old fop bug (spurious "duplicate id" error) is biting me.
I generate a .fo file like this:

$ xsltproc /usr/share/sgml/docbook/xsl-stylesheets-1.62.4/fo/docbook.xsl 
templating-dtml.dbk > templating-dtml.fo

fop chokes on page 21 out of 27 with the message:

[ERROR] file:/.../templating-dtml.fo:726:554 The id "id2450246" already exists 
in this document

That id occurs on this element:

<fo:list-item-label id="id2450246" 
end-indent="label-end()"><fo:block>Q:</fo:block></fo:list-item-label>

and it occurs once only. If I hack it out manually, fop is happy.
I've posted the above here as well:
   http://nagoya.apache.org/bugzilla/show_bug.cgi?id=14962
If anyone has enlightenment or more info, feel free to follow up
there.

I also tried with passivetex (via xmlto) which worked once
I bumped some settings:

jean@blommie courses $ grep -A1 njj /usr/share/texmf/web2c/texmf.cnf
%njj pool_size = 250000
pool_size = 500000
--
%njj save_size = 4000   % for saving values outside current group
save_size = 16000       % for saving values outside current group

Unfortunately the generated PDF doesn't cut it. The screenshots
(about 600px wide) are wider than the page, and there are TeX
artifacts ("- -4pc - -4pc", "0.60+1em") here and there.

Apart from that, both have bad widowed/orphaned lines problems,
but the output is usable if one isn't too fussy. I'm considering
using http://www.easysw.com/htmldoc/ to go from HTML->PDF.

If anyone has configuration recipes or RTFMs they'd like to share
for improved PDF (or any print format) generation from DocBook
with open source tools, feel free ..

Oooh, I've just given
   http://www.xmlmind.com/foconverter/downloadperso.shtml
a try, and it's Not Bad! The output is on par with the above
two, but it's tweakable in OpenOffice.

-- 
Jean Jordaan
http://www.upfrontsystems.co.za