Ctrl+Alt+Enter pasted while Ctrl+Alt were still down, so the target app got
Ctrl+Alt+V; accepting now closes the panel and waits for the modifiers to be
released. Candidates are accepted on pointer press because the list is
redrawn as new candidates stream in, which swallowed clicks.
The typing gate identified the focused field by its bounds, so chat boxes
that grow while typing looked like a new field on every keystroke and were
reported as "not typing". Fields are now keyed by window, control type and
name, and a mouse click re-baselines the text instead. The decision log
includes both gate values.
The live-caption model selector moves to the caption section of the General
tab, next to the other caption settings.
The speech engine now keeps an auxiliary model next to the dictation model
and transcribes with whichever the request names, reloading it once if the
engine restarted. Settings > STT gains a live-caption model so captions can
run on large-v3-turbo while dictation keeps its own model. The runtime
minimum rises to 1.7.0 because older engines would silently ignore the
model choice.
Suggestion paging moves to Up/Down: the page follows the selection and the
last item waits while more candidates are being generated. The Left/Right
page shortcuts are removed; they did nothing until a page had filled and
clash with Intel's display-rotation hotkeys.
Replaces the fixed six-second batches with a streaming track per audio
source: the uncommitted audio is re-recognised every second and sent as a
partial with its agreed (stable) prefix, a short pause finalises the line,
and long unbroken speech is committed at Whisper segment boundaries. Idle
audio is trimmed so silence cannot produce invented sentences.
Finished lines are corrected by the local model against the previous lines
and replaced in place; edits that change too much are rejected. The
behaviour can be switched off in Settings.
Next-sentence suggestions now arrive one at a time up to twelve, shown three
per page with Ctrl+Alt+Up/Down to move, Left/Right to page, Enter to accept
and Esc to close; old default bindings migrate and the panel guide follows the
live bindings. The overlay is redesigned, stays put while candidates stream
and sits outside the input box when no caret is reported.
The personal phrase memory stops learning from terminals, code editors and
the coding-agent hub, ignores symbol-heavy lines and empty-field placeholders,
and prunes existing entries that break those rules.
Fixes suggestion keys starting dictation, installs stuck on a pre-1.5.0
speech engine without the focus endpoint, Ollama runner windows flashing
while typing, the speech engine starting twice, and cold-model timeouts.
Live captions can be dragged to a remembered position and show a waiting
notice until the first line arrives.
Bumps the product version to 1.6.0 (Android/iOS build 1060000).
Adds next-sentence suggestions while typing, weekly input insights and a
personal phrase memory to the desktop app, and fixes custom instructions so
they process the text instead of inserting the instruction's own wording.
Local model requests are now bounded and individually cancellable.
Bumps the product version to 1.5.0 (Android/iOS build 1050000), refreshes the
landing and web download links, and records the new INPUT feature rows and the
open verification gaps in the infrastructure map.
Shortcuts gain several bindings per action, mouse buttons and a searchable
picker, and the stored shortcut schema changes shape, so this is a minor bump.
Existing shortcuts migrate on first launch.
Play release notes state that this release carries no mobile feature changes;
the work is desktop-only.
The Ollama setup guide still led with two retired models while the app already
defaults to the newer one, so a fresh setup would install a model the app does
not use. The guide now recommends the same model as the rest of the app and
lists current lightweight alternatives.
Installing the previous build left an app that could not update itself: the
packaging path used to guarantee the native module build does not create the
updater configuration file, so the update client had nothing to read.
That file is now written from the single feed source and its presence in the
packaged app is checked before anything is published, so an installer that
cannot update can no longer be released.
The released installer could not start: it carried a better-sqlite3 build for the
host Node runtime instead of Electron, so the app died immediately with a module
version mismatch when it opened its database.
Packaging now proves the Electron build of every runtime-sensitive native module
before an installer or archive exists, and installers are produced only from that
verified tree, so the mistake cannot pass silently. The release pipelines run the
same check.
The default local model also pointed at a retired model: a *.gguf name that
Ollama cannot serve, while the settings, onboarding, and guide screens
recommended an older model. All of them now use the model the service code
already preferred.
Auto-update could not work at all: the installer was 189 MB because it carried
the local speech engine and ffmpeg, and the download feed rejects uploads over
about 100 MiB, so update metadata could never be published.
The installer now leaves those components out and the app fetches them the first
time they are needed, verifying every part and the joined archive before
installing. The installer is 90.6 MiB, the update feed is published again, and
updates stay small because the engine is not re-sent on every release.
The fetch is visible and recoverable: the download runs with progress, a failed
install cleans up after itself, and Settings > STT shows the runtime status with
a manual download action for when the automatic one cannot run.
The manual install path still needed 7-Zip, which the target machine does not
have, so "installable without a certificate" was not yet true.
The channel now also publishes the app as byte-split zip parts, and the install
script joins them and extracts with the built-in Windows Expand-Archive after
verifying every part and the joined archive. Version 1.3.1 republishes the
channel from a single build, because a version's artifacts can only match one
build and published volumes are never overwritten.
Installers could not be published at all: the signing certificate does not
exist yet, and the release pipelines stop at their signing guard. Users had no
way to install a fixed build, so the product was effectively stuck behind a
certificate that takes weeks to obtain.
There is also a second, independent blocker: the download feed sits behind
Cloudflare, which rejects any upload body over about 100 MiB, and the app with
its speech engine exceeds that even when signed.
A portable channel now publishes what can actually be delivered today: the app
compressed into 95 MiB 7z volumes (162 MiB total instead of 243 MiB), a Scoop
bucket for a normal install and uninstall experience, and a verifiable manual
installer script. It is deliberately separate from the auto-update feed, needs
no certificate, and refuses to overwrite an already published version.
Local dictation had never produced a transcript on an installed build. The
engine itself was healthy; every connection to it was broken.
Installed builds shipped no speech engine at all: the packaging config had no
entry for the faster-whisper sidecar and no pipeline step built one, so the app
always fell back to a system Python without the runtime. Development was broken
too, because the sidecar and SoX paths were resolved against the Vite output
directory instead of the app root, which also meant recording failed with a SoX
ENOENT. On hosts where localhost resolves only to IPv6, every local request was
refused outright, which silently disabled both local transcription and the local
LLM.
The sidecar is now built and bundled (including the Silero VAD data it needs),
gated by a packaging check that fails when the engine or its data is missing.
Paths are discovered from the app root and fail loudly when the engine is
absent. Local engine URLs are normalized to the IPv4 loopback, decoding is tuned
so repeated hallucinations cannot compound (the same transcript now takes about
a fifth of the time), the engine is warmed up at startup, and holding the hotkey
now shows the text forming live in the recording tip.
The changelog still described unreleased work under 1.1.0, which was already
published with its own notes. Those notes are restored verbatim for history,
and the new work has its own 1.2.0 section that the feed publisher will turn
into release notes.
The release guide, infrastructure map, and mobile SSOT now carry the 1.2.0
identity, state that installer binaries are distributed through the feed and
never committed, and record that the published 1.1.0 installer is unsigned and
is being superseded rather than rewritten. Backlog entries cover the remaining
external signing and token secrets.
Release notes for 1.1.0 were split between an Unreleased section and the
version section, so the published notes would have omitted the update-feed
and desktop changes. Everything shipping in this version now sits under one
`## [1.1.0]` heading.
`docs/map/` becomes the entry point for what infrastructure exists per
platform and how far each feature is developed, with a documented update
protocol so feature work and this map do not drift apart again. The release
guide now states that installer binaries live in the update feed rather than
the repository.
refactor-wave WS2가 AppConfig 인터페이스에 7키 추가 + as never 16건 제거했으나
CONFIG_DEFAULTS(ConfigService.ts) 기본값 누락 → 기존 사용자 config(0.1.x) 마이그레이션
시 configGet undefined → main 프로세스 .map() 크래시 (v0.2.0-alpha 앱 실행 불가).
수정: CONFIG_DEFAULTS에 7키 기본값 추가.
검증: dev 실행 — main 빌드 + electron start 정상 (수정 전엔 이 단계 크래시).
교훈: 배포 전 dev 실행 + 산물 e2e 필수. typecheck GREEN ≠ 런타임 안전.
- CHANGELOG.md: Phase 1~13 + 랜딩 페이지 전체 변경 이력
- LICENSE: MIT
- CONTRIBUTING.md: 개발 환경, 코드 표준, PR 프로세스
- SECURITY.md: 보안 정책, 취약점 보고 절차
- .editorconfig: 에디터 설정 통일 (indent, charset, eol)
- .nvmrc: Node.js 20
- .gitattributes: line ending 정규화, 바이너리 선언, linguist 설정
- .gitignore: site/dist, Python, Whisper 모델 등 누락 항목 추가
