# OpenSanta Project roadmap

Public plan and timing. The technical training detail lives in [TRAINING-PLAN.md](TRAINING-PLAN.md). Status: **Formation**. Nothing has been posted, and submissions are not open.

The website progress bar is driven by [`status.json`](status.json). See the README for how to update it.

## Stages

1. **Formation** (current) — Establish the project, its ideals, consent process, docs, and public home. Create the project email and both forms before any public ask.
2. **Recording** — Volunteers sign consent and record the shared script (core set plus one pack).
3. **Dataset** — Review, quality-check, and publish consented recordings under CC BY 4.0.
4. **Training** — Train a high-quality blended classic American Santa voice. Compute waits until the recordings are in.
5. **Release** — Release the free, open Santa voice for anyone to use, for any lawful purpose, including commercially.

## Timeline

**Now (mid-November 2026) — awareness and formation.**
Build the public home and finish drafts. Before any public post: the project-owned email, the no-upload sign-up form, and the Google upload form need to exist, and the placeholders in the public drafts need to be replaced. A lawyer's review of the consent form and licenses is recommended before taking real submissions.

**December 15–24, 2026 — go easy.**
This is many Santa performers' busiest week. Don't chase people then. A quiet listing of the sign-up link is enough, once it exists.

**December 26, 2026 through January 2027 — main recording window.**
The natural time to record, while the season is still fresh and people have more time. First milestone: about **1 hour of accepted audio, from about 12–15 people**, with every pack covered by at least two speakers. That's the point the training plan gives for starting training experiments.

**Late winter into spring 2027 — toward a good v1.**
Keep accepting careful, consented recordings rather than rushing. v1 milestone: about **2–4 hours of accepted audio, from about 25–50 people**. A smaller on-device voice can come later.

## Still open

- The [model license](LICENSE-MODEL.md). OpenRAIL-M style is recommended, with Apache-2.0 as a fallback.
- Project email set up: support@opensanta.org.
- Setting up the project Google account, both forms, and the private Drive archive.
- A lawyer's review before launch, including consent and biometric-privacy questions.

Training, GPU credits, and a smaller on-device voice all wait until the recordings are in.
