I Just got back from the hospital, where I spent a few days recuperating from a nasty bug.
Congrats on the release of 0.6 stable. Coming back to my last complaint, calibre still has problems converting multi-page html books, like this one:
http://www.freenetpages.co.uk/hp/alan.gauld/tutor.zip
I tried your suggestion of unzipping the book and pointing calibre to the contents file (in this case tutcont.htm), but this only generated new problems, without completely fixing the previous ones. So now:
1. The calibre generated zip file contains corrupted versions of some of the pages (why does it fiddle with them in the first place ?), e.g. tutseq1.htm
2. The order of the pages is still not quite right. Some sections still get mixed up, despite having pointed calibre to the contents page that has the correct order (e.g. the "getting started" and "simple sequences" chapters get reversed for some reason).
3. Despite pointing calibre right at the contents page, it does not recognize it as such and generates its own contents page, that, simply put, differs radically from the pre-existing (and obviously correct and desired) one.
-As a side note... what should the name of the table of contents page be for calibre to recognize it as such and not generate a new one?