Have been spending quite a bit of time reading e-books lately, and the quality is pretty horrendous overall.
Main errors:
Typographical errors
Layout/formatting glitches
Lack of font uniformity
Poor handling of illustrations and "special characters" such as math symbols.
Main reasons for errors:
Widespread use of OCR (supposed "smart" character recognition within a scanned text) without subsequent human proofreaders
Conversion mistakes (changing one electronic format to another)
Lack of uniformity in fonts and word-processing applications
Change of corporate culture from giving-a-shit to laying-off-and-praying
I'm trying mostly to read .epub books and avoid Amazon/Kindle but occasionally I still have to resort to Kindle/mobi/azw and the situation is no better. Surprisingly, I've found public domain works from Feedbooks.com to be of better quality than many offerings from "respectable" mainstream publishers, although there are no guarantees.
This will be a series of blog posts (I hope) that document egregious e-book errors.
Let's start with a doozy, from Richard Ellman's Yeats, The Man and the Masks, 1948, e-book (c) 2016 by Pickle Partners Publishing. Ellmann is discussing an early draft of the poem that became "To his Heart, bidding that it have no Fear" (1896):
Astonishing, indeed. The word is supposed to be "part" -- only a demon would introduce such an error into such a lovely poem.
Also, note the weirdly italicized third line -- the text is supposed to be in italics from that line to the end. Pickle Partners, get thee hence to the typesetters.