prosolis
77f284f65c
`pair_lang` had always been answering a second question nobody asked: it says which two languages, and every surface built on it assumed English was the one being learned. That is why hanzi is never tokenized, never spell-checked, never glossed — correct for a Mandarin native practising English, backwards for an English native practising Mandarin. `users.direction` (migration 0016) separates the two questions; a `zh-learner` pair code would have been cheaper and would have made two directions of one pair look like two unrelated languages to every query. Segmentation is what replaces `wordAt` where there are no spaces: a shortest-path walk over log-probabilities, 232 ms and 14 MB for 188,522 words. The browser gets the word list because segmentation runs on hover; the server keeps the whole dictionary. Their coverage gates come out opposite on purpose — the client list is frequency-gated because the segmentation is measurably identical without the tail, and the dictionary is gated by nothing, because its only power is to explain and the word a learner stops on is the rare one. The 错别字 pack is 24 confusable pairs behind two mechanical gates. One admits a pair only if the wrong form is not a dictionary word and the right form is, which is why it refuses 自已 for 自己 — a real error whose wrong form is a headword. The other asks the segmenter whether the two characters already belong to two different words, without which 自己经常, 睡觉的时候 and 不知到底 would all be corrupted silently into text still made of real characters. Not deployed (this carries a migration), not seen in a browser, and no account has ever been in the learner direction. The IME composition guards were in scope and are not done — see BUILD_PLAN Phase 26.
🌸 Petal
A self-hosted, privacy-first writing editor with warm bubbly design, auto-save, and
local-LLM grammar/ESL suggestions. See petal-spec.md for the full
design spec and BUILD_PLAN.md for build progress.
Stack
Go + chi backend · SQLite (modernc, pure Go) · React + Vite + Tiptap + Tailwind v4 frontend ·
local vLLM/Ollama for AI suggestions. Single-binary deployment (frontend embedded via go:embed).
Local development
Two processes during development:
# 1. Backend (serves /api on :8080)
go run ./cmd/server
# 2. Frontend dev server (HMR on :5173, proxies /api → :8080)
cd web && npm install && npm run dev
Open http://localhost:5173 while developing.
Production build (single binary)
cd web && npm run build # emits web/dist (embedded by the Go binary)
cd .. && go build -o petal ./cmd/server
./petal # serves UI + API on :8080
Configuration is via environment variables — copy .env.example to .env.
Deployment
docker compose up -d --build # petal + the two Piper read-aloud sidecars
Behind Traefik on the parodia.dev VPS; vLLM stays on millenia over headscale.
Full runbook — first deploy, the LLM link, backups and restore — in
deploy/README.md.
Status
Early build, multi-session. Auth (Authentik OIDC) is next; Copyleaks plagiarism is
still parked — see BUILD_PLAN.md.