Based on the info above, I tried PDF2HTMLgui. I used it on the Pimlico Datebook5 manual, which is fairly complicated, and has lots of pictures with text around them (screenshots mostly).
It works great using the method I describe in what follows. But as you can see at the bottom of this post, I can't convert it into a nice iSilo document.
Any ideas? (Hope I didn't repeat too much from the rest of the thread... I did read it but too quickly.)
The download I used was at: http://guiguy.wminds.com/downloads/p...lgui/down.html
The idea is that this program is just a gui shell, and the guts are contained in two additional required programs (as indicated on that page) that must also be downloaded:
GhostScript (v8.14) and PDFtoHTML
You need to be sure to grab the windows binary executables, not the source files.
To install, you install ghostscript, and extract PDFtoHTML files into a directory of your choice (that's the command line program). Then extract the gui program .exe into the same directory as the command line program.
When you run it, you'll need to tell it where to find the ghostscript executable. There were two to choose from, and I took the first alphabetically. I think it was without the "c", but I can't remember.
You then give it the source file .pdf location (make sure the directory names are not too long because it will choke if the command line is more than 255 chars, which includes the full path and filename of the source and destination, plus more).
It supplies a candidate destination. You can choose a bunch of options I don't understand, but the key option is the "generate complex document" which is visible as a conversion checkbox after you choose "more options". When I didn't choose it, the conversion was just text and didn't accomplish much for me at all.
The complex option gives me a "framed" page HTML result with index on left side and single pages on the right side, and it looks gorgeous. You might want to choose another folder for the result because there can be a LOT of files.
The conversion went pretty quickly, and was very effective. Note also that the complex option requires the ghostscript executable.
THE REMAINING PROBLEM...
I haven't figured out the remaining trick of getting it into iSilo... when I do a straight conversion of the index page, I thought I'd get a nice document. But all I got was pages with the pictures and separate text. Not pictures on a page with text surrounding it at the right location.
Interestingly, it opens and displays fine with Internet Explorer, but with MS Word, it shows improperly just like in the converted iSilo document.
One would think that the HTML conversion works just like any other iSiloX conversion with images (after all that work to get images in the document, you probably don't want to lose them now!) And I always convert to a memory card destination. But the results were not right.