Postr 1.1.1 keeps dictation in the composer

Postr began with a severe interface rule: writing and publishing a note should feel like one action. Version 1.1.1 extends that rule to speech. Tap the microphone in the top bar and the composer remains in place while Postr listens. The main button area becomes three choices: Cancel, Paste, and Post.

Those choices remove the old procession of screens. Paste puts the transcript into the draft at the position captured when dictation began. Post combines that transcript with the same captured draft and waits for an explicit tap before signing anything. Cancel stops the session without changing the note. The transcript stays visible inside Postr during review, beside the controls that decide its fate.

Android speech providers do not all expose the same interface. When the installed provider offers a compatible recognition service, Postr opens an app-owned microphone session and sends the audio through that service without handing the interface to a provider activity. Postr checks the selected service, installed package version, permission, and capture route again at startup. If Android exposes only an activity, Postr can still use that provider fallback instead of pretending an in-app session exists.

Mentions received the same treatment. Search follows the caret, including an @name in the middle of existing text, and inserting @ immediately before a written username starts the lookup. Choosing a profile turns the visible name into a structured mention whose color, background, and weight differ from ordinary text. An unselected @name remains plain text, so the composer shows which names will become Nostr references before Amber signs the note.

Postr still asks Amber for the same narrow signing capabilities, and dictation never turns speech into a signature by itself. Version 1.1.1 is available through Zapstore for anyone who wants fewer screens between a thought and a verified post.