The book that won’t open

An EPUB is a small website in a zip

Rename an .epub to .zip and open it: inside are chapters written as web pages, the pictures, a stylesheet, and two short lists — the order the chapters are read in and the table of contents. That is the whole format. It also means the most natural program to read one is the one you already have open: a browser has been rendering pages like these for thirty years.

Yet Windows has no EPUB reader of its own — the old Edge could open them, the current one cannot. An iPhone hands the file to Apple Books, and Android usually wants an app first. So our EPUB page does the simple thing: you choose the file, the tab unpacks it in memory, and the book opens with its own chapter list. Nothing is uploaded; there is nowhere for it to go.

FB2 and the page of question marks

FB2 is the other common book file, and it lives almost entirely in Russian-language libraries. It is a single XML file with the pictures packed inside as text, often zipped into .fb2.zip. No operating system opens it without an extra app.

The trap with FB2 is age. A large share of the files in circulation were made before everything switched to UTF-8 and are stored in windows-1251 or KOI8-R. A reader that assumes UTF-8 shows you a page of “����” instead of a novel. The first line of an FB2 says which encoding it uses, so the FB2 page reads that line before anything else — a book from a decade-old library opens as text.

Three files we refuse, and say why

Some books cannot be opened by anyone except the shop that sold them, and pretending otherwise wastes your evening. So the reader checks the bytes, not just the file name, and answers plainly:

• DRM. Books bought from Kindle and most stores are encrypted for the shop’s own app. We will not try to break that, and the reader tells you so instead of failing quietly.
• MOBI and AZW3. These are Kindle formats. A file renamed to .epub is still recognised for what it is — and you get a sentence, not a crash.
• A PDF dropped in by mistake. It is pointed to the PDF tools, which is where it belongs.

The mistake we made first

An EPUB can carry a file called encryption.xml, and DRM uses it. So the first version of the reader treated that file as proof of DRM. It is not: many publishers use the same file only to scramble the fonts embedded in the book, so that the font cannot be lifted out and reused. The text is perfectly readable.

The result was a reader telling people that a book they had bought DRM-free could only be opened by the store’s app — a false statement, aimed at exactly the person the page promised to help. It now reads what that file actually lists, opens books where only fonts are scrambled, and keeps the refusal for books whose text is locked.

What is remembered, and where

On a tool page the book opens and you read, chapter by chapter. Open it on the desktop and the reader keeps more: your place in the book, serif or sans, a text size from 15 to 26, a light, sepia or dark page, and your bookmarks. You can search the whole book, and a passage you select can be saved as a quote into Notes, with the book’s name on it.

All of that lives in this browser on this device. There is no account, which means there is no sync: the phone does not know where you stopped on the laptop. The reader says “stored only on this device” next to the bookmarks, because the alternative — letting you find out after clearing your browser — is worse.

What it is not

It is not a bookshop or a library. We do not host books, look them up or link to them; the reader opens the file you already have. It does not open Kindle books, and it does not sync. If you read every night across three devices, a dedicated reading app will serve you better, and we would rather say that here than have you discover it on page two hundred.

If what you have is one file that will not open, it opens in the tab you are already in.