Originally Posted by rlauzon
But pay sites, like Fictionwise, also have a good number of errors for longer eBooks. One I downloaded recently had typical OCR errors in them (ex: the main character's name is "Varl", but it often showed up in the text as "Van" - the OCR software mistaking "rl" for "n").
That surprises me. Shouldn't they be able to work with the original electronic transcript instead of first having to OCR it?