Elf Sternberg

Jeez, Elf — to be blunt: you're a software engineer, think like one. If you were building their system, and you had to construct the behavior you've just described, how would you do it? My first guess would be that they keep a db of book-and-page, and that when you flip a page the info about what page you're on is sync'd up to their server. For books on their store the key is presumably the backend ID; for sideloaded works where they don't have an ID it's probably something like title or title-and-author, though filename is another reasonable possibility. Why would they waste bandwidth and disk space on transferring some sideloaded behemoth onto their servers? They certainly know what the filenames on your device are (and probably whatever the book equivalent of an ID3 tag is), but I honestly don't see the business case for taking more than that unless they're trying to be actively evil.