Skip to content

v1.2.0: whole documents, resume, maximized reader, store-ready

Choose a tag to compare

@CryptoJones CryptoJones released this 13 Sep 06:50
· 22 commits to main since this release
0dbb3d9

Read entire documents, and pick up where you left off

  • Read entire document: Alt+R with nothing selected, Read entire document with Emma in the context menu, or the popup button reads the page's main content (Mozilla Readability, vendored). The title is spoken first and shown in the header.
  • Resume: a document you stop partway remembers its place per URL. Reading it again resumes there; the popup offers Resume reading document · n / N, and Read entire document starts over. Finishing clears the bookmark.
  • Maximize: ⤢ in the header or Alt+M fills the tab with the word window; Esc restores. Remembered across reads.

Under the hood

  • Server audio cache bounded per job (a whole document no longer costs hundreds of MB); a bad sentence is skipped instead of ending the read.
  • Server answers cross-origin requests from extension origins only. Restart the server to get this: systemctl --user restart clarkreader or rerun server/run.sh.
  • Store readiness: minimum Chrome 116, optional host permissions for a server on another machine (the Chromebook case), a setup guide on first install and from every "cannot reach the server" error, permission justifications and privacy policy in the README.

Install

The zips below are the store packages: unzip and load unpacked (Chrome: chrome://extensions with Developer mode; Firefox: about:debugging, temporary add-on). The server must be installed and running; the extension opens the guide on first install.

25 extension tests, 37 server tests. See #4.