| # Plan: talk — Reachy Mini date/time announcer |
|
|
| ## What I understand you want |
|
|
| A Reachy Mini Python app (`talk`) that: |
| 1. Wakes the robot up (animation + sound) |
| 2. Speaks the current date and time in German via text-to-speech |
|
|
| ## App flavour |
|
|
| **Python app** (not JS) because: |
| - TTS requires `espeak-ng` on the robot |
| - This is a developer-style on-robot app, same pattern as the sibling `recognizer` |
| - The `--template conversation` is designed exactly for "making the robot talk" |
|
|
| Scaffold command (after plan approval): |
| ```bash |
| reachy-mini-app-assistant create --template conversation talk /Users/oliver/Documents/dev/reachy_mini/talk --publish |
| ``` |
|
|
| ## Technical approach |
|
|
| 1. **TTS**: `espeak-ng -v de -s 130 -w /tmp/datetime.wav` + `reachy_mini.media.play_sound()` |
| German voice gives natural German phonetics for date/time strings. |
|
|
| 2. **Date/time string**: Python `datetime` formatted manually in German |
| e.g. `"Heute ist Montag, der neunzehnte Mai 2026. Es ist 14 Uhr 30."` — spelled-out numbers sound better from a TTS engine than digits. |
|
|
| 3. **Lifecycle**: `ReachyMiniApp` subclass, `--template conversation` base |
| - On `run()`: call `reachy_mini.wake_up()`, then speak the date/time string |
|
|
| ## Open questions — please fill in your answers |
|
|
| **Q1: Trigger** → B) Antenna press (either antenna → speak) |
| **Q2: Hardware** → Wireless (onboard CM4) |
| **Q3: After speaking** → Goes back to sleep |
|
|
