Print, yes, But they have to convert to ebook files, which have to discard a lot of PDF information. Some people may think it easier just to extract the text.
I also believe that, before ebooks, there was a general practice of discarding all digital files but the one specifically used to print, and that file was not text but instructions on how to draw said text. Think of it akin to taking a screenshot of a graphics file and then destroying the original photoshop. And even if you didn’t destroy it, you’d store it on, say, a floppy and it would deteriorate.
And even if the original author still had the file, how often do you think the publisher contacts the author and gets the file from them, rather than just using what they have to produce the ebook? As said above, many people don’t even get the difference.
I also wonder if you could combine crowd sourcing with software that would check and see if it’s possible that the original did say what the crowd thinks it said, perhaps even letting the crowd pick from multiple possibilities, like how Google Translate works.
I’m reading an ebook I got from Baen, and have noticed it’s got a lot of errors as well. Missing spaces, extraneous mid-word spaces, the same sort of letter-mungling the OP cited, etc. Really annoying.
That would explain it, but where did you get that information? No self-respecting IT department would ever recommend that, no sirree. The print file is the most expendable, as it can be recreated as long as you have the original app and original data.
As BigT noted, an e-book file is not a PDF file. It’s a different format. So some kind of translation is required, if you’re starting with a PDF and ending with an e-book.
Two hours ago, I was proofreading a book by a well-known author and the copyeditor didn’t notice the author’s alluding to Oswald’s having shot Ruby (the name transposition was unambiguously unintentional).
I encounter ludicrous stuff like this constantly now, in addition to the formatting/mechanical errors I alluded to previously, such as commas that s/be periods and vice versa, that were relatively rare before the advent of electronic copyediting. This sloppiness on the part of most copyeditors nowadays was definitely not the case, say, 5 years ago, when I didn’t have be constantly wearing both proofreader and copyeditor hats. More and more, careful and conscientious “line editing” appears to have gone the way of the quill pen.
I am fully in agreement with your generally positive take on freelancing in general (my last 9-to-5 job ended in 1982) …
I know that. My point is you would never start with a PDF if you could help it; no-one writes and designs books for printing directly in PDFish. Typically they are written and edited in Word, then imported into InDesign (or conceivably QuarkExpress) where they are designed and re-edited. Once done the InDesign files are exported into PDF for printing, and in parallel fed into the e-publishing process to make ePub, mobi or whatever formats you are publishing in.
It would only be if you had lost the pre-PDF files (or the book is so old that it never existed in soft form) that you’d need to go from PDF -> ePub, or scan the hard copy.