robert said:

Grabbing only text is seldom worth it.. nothing to compare it against.

um, i would disagree. it is your job to create something to compare it against. if you can't find page-scans, then do the scans yourself, and run those scans through distributed proofreaders. then compare the d.p. output against that other text. (because this is the goal, you'll want to make sure the p-book you scan is the same edition as the other text.) this is absolutely the best way to obtain accurate text, by comparing two independently-produced versions... and yes, i know y'all know that. as we speak, even, a version of "the secret garden" is going through this. it just seemed like this little reminder was being called. -bowerbird