Release log — DogEar

Every build of this app, what changed, and where its release checks stand. We publish the state of a build honestly — including while checks are still running.

2026-10-16 — gate results, and a speech fix

Our release gate has checked the three parts of this launch. Its verdicts, in its own terms:

reader-core v0.2.0 (the reading engine) — PASS WITH FINDINGS.
This web app — BLOCKED at the gate on a silent speech failure, described below. Fix in loop; re-gate to follow.
The installable app (PWA) — PASS WITH FINDINGS; install and offline proofs outstanding.

The block: in a browser that offers the speech feature but has no working voice, pressing Listen produced no sound and no warning — it failed silently. This build fixes that: if your browser has no usable voice, the app says so up front, in both the library and the reading view; and if a Listen press produces no speech within about two seconds, playback stops and the warning appears instead of silence. The link importer also now refuses two more spellings of private network addresses that it previously left to the platform to refuse, and a broken link in the app’s footer is fixed. The gate re-checks this build before its verdict can change.

2026-10-16 — the web app, first build

Status: release candidate. This build is live and in its one-week candidate window under review by our release gate. Gate verdicts: recorded in the entry above.

The DogEar web app: add a document by pasting text, uploading a .txt/.md/.html file, or fetching a public https link; hear it read aloud; your place is kept on this device and offered as a resume point when you return.
Installable as an app (PWA): add it to your home screen and it opens and reads your saved documents offline.
Reading engine: reader-core v0.2.0, pinned. That engine release carries the place-keeping, parser-order and bare-text fixes the site embed earned in production, and prefers on-device voices.
Free. No account. Documents stay on the device (see Privacy).