There seems to be some kind of error when converting corpus-content.xml to the forrest format. The "forrest build" stops about halfway through the generation of the website.
The end of the xtdoc/sd/build/webapp/WEB-INF/logs/error.log contains a lot of sentences like these:
ERROR (2006-06-15) 12:22.22:854 [cocoon.manager.fop] (Unknown-URI) Unknown-Thread/MessageHand
ler: Error in number-rows-spanned property value '': org.apache.fop.fo.expr.PropertyException: No
I don't know what is wrong, corpus-content.xml validates and looks like it's supposed to. could the problem be in the interaction of the xml-structure and Sjur's scripts that create the content of the page?
I have now updated the corpus summary processing to only generate a fraction of the file size compared to earlier. Please 'cvs up', and rerun Forrest to see how it behaves. 'forrest run' is now processing the sme corpus content listing in about 2 seconds on my portable, which seems promising to me.
It seems to work perfectly now.
Accidentally marked this bug as fixed.
erh, this was fixed, after all ... sorry.