I can confirm the 100 char limit on the xml file. At least that happens on mine.
However, the reader seems to have some information on the full annotation in the <start> and <end> tag. If you decode them (in
http://www.opinionatedgeek.com/dotne...e/Default.aspx for instance) they look like:
#pdfloc(b835,3,1658,0,0,0,0,1)
I wonder if there is any tool that can make sense of that...
Best