A reading app should not phone home to remember your page.

Most e-reading ecosystems treat your library as a hostage. Upload to the cloud, accept the terms, hope the service does not change. The app is nice until it is not.

I wanted an e-reader that treated a book like a file and my reading history like local data.

Kindle and Apple Books are convenient until you want to leave. Your annotations are locked inside their sync service, and their apps do not treat the book file as something you own.

Varq is the opposite: the file stays in a folder you control, and the annotations are keyed to that file. You can switch apps without losing your reading history.

Varq reads from a folder you pick. When you import a book, the app copies it into a directory you control and builds a SwiftData index for search, progress, and recent items. Highlights and notes are stored as annotations tied to the file hash, so they survive app reinstalls as long as the book file survives.

EPUB and CBZ parsing happens locally. There is no cloud conversion, no analytics ping, and no “enhanced reading experience” that phones home. The PDF renderer uses Apple’s native PDFKit for performance and to avoid shipping a third-party binary that could exfiltrate pages.

SwiftData handles the metadata layer. I versioned the schema from the start because Core Data migrations taught me that schema drift is painful.

Varq is not faster at rendering a page than a web wrapper, but it is faster at everything around reading: opening the library, resuming the last position, searching highlights, and switching books.

Local-first apps force you to design for durability from day one. If a user loses their data because they deleted the app, that is your fault, not Apple’s.

Native UI matters for reading. A web wrapper can display text, but page-turn physics, selection handles, and annotation margins are the reasons people stay.