I understand seeing typos in books derived from the darknet, where pbooks have been scanned and OCRed. OCR is not an exact science and errors inevitably crop up, but I don't understand why there are typos in ebooks from publishers.
After all, they already have electronic versions of the books that feed the typesetting software. Even the most entry level programmer should be able to produce a script to convert that to any ebook format necessary. So why do we still get typos??
