Keysat
0ae59f3550
Introduces RECAP_MODE=multi alongside single-mode self-host: - Tenant auth + accounts (magic-link via System SMTP), per-tenant credit pool, anonymous trial minting with per-IP/-64 caps - Self-serve Pro/Max purchase: inline Lightning (BTCPay) + card (Zaprite), prepaid 30-day periods, expiry-reminder emails - Core-decoupling: relay owns cloud tier/expiry keyed by Recaps user-id - SQLite (better-sqlite3) schema for multi-mode; filesystem unchanged for single - StartOS actions/versions through 0.2.155
14 lines
1.1 KiB
TypeScript
14 lines
1.1 KiB
TypeScript
import { VersionInfo } from '@start9labs/start-sdk'
|
|
|
|
export const v_0_2_131 = VersionInfo.of({
|
|
version: '0.2.131:0',
|
|
releaseNotes: {
|
|
en_US:
|
|
"Frontend half of relay v0.2.100's small-cluster suppression. Two new chip states: (1) UNCERTAIN — the post-cluster suppression pass reassigned a small cluster to one of the main anchor speakers as best-guess attribution. Chip renders with a '?' suffix (e.g. 'MH?') and the tooltip notes 'best-guess attribution'. (2) UNKNOWN — the special Speaker_Unknown pseudo-speaker grouping brief utterances that didn't confidently match any main anchor. Chip is grey ('?'), legend reads 'Unknown', and it sorts to the end of the legend after named speakers. The chip color and class for any given speaker stays stable as before — visual identity is preserved across the transcript. Existing low-confidence rendering (per-segment diarize confidence < 0.5) continues to work; uncertain + low-conf now BOTH trigger the '?' suffix.",
|
|
},
|
|
migrations: {
|
|
up: async ({ effects }) => {},
|
|
down: async ({ effects }) => {},
|
|
},
|
|
})
|