A one-time refresh, and why

The next update — landing in the coming days — brings a change you'll hear immediately: Toki Pona and Esperanto now speak entirely through FlashBoss's native neural voice, the one trained specifically for Esperanto's sounds. Toki Pona had been borrowing a German voice, and it showed on the smallest words: vowels blurred, and w drifted toward v — sounds that voice was never taught. Every one of Toki Pona's fourteen sounds is native to its new home. Esperanto players gain too: any clips left over from its earlier borrowed voice are retired, so everything you hear is the native voice, consistently.

To make that true on every machine, the first launch after the update performs a one-time cleanup of the cached audio for these two packs. You don't need to do anything: audio regenerates in the new voice as you play, or in one pass through the Audio Cache Manager. Your saves, progress, streaks and settings are untouched — this only affects generated sound files.

What else rides along

• Save reliability hardened end to end — your progress is secured on your machine first, card by card, and the cloud copy now catches itself up at both ends of every session.
• The Audio Cache Manager was rebuilt on the same engine as live speech — its coverage counts are now honest, and caching works for every installed pack.
• Lessons across the whole library were trimmed to fit their frames — no more pages running past the box.
• For Windows players, a fix is on its way that stops the game working your machine while it waits for a keypress — quieter fans, longer batteries.
• And the new voice architecture opens a path we intend to walk: more native voices, and real voice choices, ahead.

The small announcement before the big one

This is the housekeeping note. The update carrying these changes is the biggest in FlashBoss's history, and it deserves its own story — the next announcement tells it in full. Watch this space.

Master Odiin