The system balloon carries the OS notification sound, which does not belong to
this app, and it only fired while the window was out of focus — with the app
focused nothing appeared at all.
Notifications now surface inside the window, carry the same synthesised timbre
as the rest of the app's sounds, and clicking one opens the channel. The native
balloon is kept for when the window is not visible, since an in-app card
nobody can see is no notification, but it is now silent: the app plays its own
effect instead.
A focused window is notified only about other channels — announcing the
conversation someone is already reading is noise.
Also fixes the Gitea publish cleanup, which silently deleted nothing: it
interpolated an Actions expression inside a bash , and when the pattern
did not match, the loop passed over every asset. The release ended with two
latest.yml files and the updater served the older one, reporting 1.1.0 as
current — an update that exists but is never offered, with no error anywhere.
The filter is plain bash now, logs what it found, and the job fails if more
than one latest.yml survives.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The main process already showed a system notification when an update finished
downloading, and nothing in the renderer listened — so the only cue vanished on
its own and was missed by anyone not watching the desktop at that moment.
A prompt now stays in the window until it is acted on or dismissed, with the
restart wired to the install-update channel the preload already exposed.
Dismissing is 'not now', not 'never': a later version reopens the prompt, since
the version it refers to has changed.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Completes the feature: the tables existed but nothing could be put in them.
Space settings gain an Emojis & Stickers panel behind MANAGE_SPACE, with a
512KB ceiling — both are fetched on every message that uses them, so weight
matters more than fidelity. The suggested name is pre-normalised so the common
case needs no typing, and a name collision reports itself distinctly from an
upload failure: the corrective action is different.
Custom emojis join the emoji picker as their own category. They have no native
character, so selecting one inserts :name: — the same text the renderer
resolves back to an image, which also means copying a message yields something
that still reads.
Stickers get a picker tab that only appears inside a space, since that is where
they exist, and send immediately on click: a sticker is the whole message, so
parking it in the composer to await Enter would make no sense.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Emojis and stickers get separate tables on purpose: an emoji sits inside a
sentence and a sticker is the whole message, so their sizes, limits and render
paths differ.
Emoji names are unique per space and restricted to letters, digits and
underscore — :name: has to resolve to one image, and a space or colon inside
the name would make the reference impossible to delimit. Sticker names are
labels picked from a grid, so they accept anything.
Rendering reuses the existing markdown pipeline by rewriting :name: into image
syntax, which needs no plugin. Code spans and fences are left alone: text
someone wrapped in backticks was meant to be shown literally, and turning it
into an image would destroy what they wrote.
Messages carry stickerId with ON DELETE set null — removing a sticker from the
space must not delete the history of everyone who used it. The server refuses a
sticker from another space, which would otherwise leak images between unrelated
servers.
Expressions load once per space and stay in memory: the message renderer
consults the map for every :name: it finds, and a request per lookup would turn
each message into a cascade.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The filters themselves already existed end to end — the server takes q, from,
has, before and after, the API client passes them, and the popover has fields
for each. What it lacked was discovery: the panel sits behind a button, so the
capability was invisible.
Typing 'de:fulano' or 'has:image' straight into the search box now applies the
same filters. Keys are accepted in both languages, since the app is bilingual,
and an unrecognised token falls back to being search text — otherwise a message
containing a URL or 'algo:coisa' would become unfindable.
Inline filters win over the panel's: whoever just typed one is expressing the
more recent intent.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Pinned state lives on the message rather than a join table: a message is
pinned in exactly one channel, its own, so a separate table would add a join to
every lookup and buy nothing.
Every message now carries its pin state, so the timeline can mark a pin without
a second request and the panel and the timeline cannot disagree. Toggling is
deliberately not optimistic — the server refuses past the channel's limit, and
showing it pinned before confirmation would lie in exactly that case.
The pins list reuses the same assembly the channel history uses, extracted into
one helper, so the two cannot drift apart in what they include.
The migration also adds the (channel, user, created) index the filtered search
will need, since both touch the same table and one migration is cheaper than
two.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Clears the two remaining rounds owed under the fork's per-update translation
rule.
Covers channels, direct messages, the context menus, the inline voice settings
and the member list's loading and grouping labels.
The member grouping needed care: one function returns either a fixed label or
a role's own name, and role names are user data that must never be translated.
It now marks only the fixed cases and the component resolves them at render,
leaving role names untouched. The function is not a component, so it cannot
call the hook itself.
Catching up on the fork's rule that every update also translates one existing
system — three updates went by without one.
The composer's placeholder takes the channel as a parameter rather than being
concatenated, since Portuguese puts the channel elsewhere in the sentence than
English does.
The translation hook is bound to `tr` in these two files, not `t`: both already
use `t` for transfer records in inner scopes, which shadowed the function and
made it uncallable exactly where it was needed.
Listeners had to run the participant at 200% with attenuation off. Two causes,
both on the publishing side.
WASAPI returns the process mix at a level well below what a microphone track
arrives at, so the track was quiet before it ever left the machine. A gain
stage now boosts it, followed by a limiter — a compressor with a high ratio and
fast attack — so the boost cannot clip loud passages.
And the track was published with LiveKit's defaults, which are tuned for
speech: mono, low bitrate, and DTX enabled, which stops transmitting during
silence. On game and music audio that reads as thin and clipped. It now
publishes with the music stereo preset, stereo forced, and DTX and RED off.
Lives in packages/web, so it reaches the desktop app through a normal deploy —
no new installer.
The originals were all 1.14s — long enough to be intrusive for events that
fire whenever anyone joins, leaves, or watches a stream.
Synthesised rather than sourced: no third-party file, so no licensing question
in a public repository. The parameters are measured, not guessed — envelope,
spectral peaks and decay taken from two reference sounds the instance owner
supplied. The timbre is fundamental plus octave at near-equal strength
(1.00 / 0.85 / 0.10 / 0.02) decaying to 1/e in 0.19s, with no reverb.
Call join rises C4→G4, leave falls D4→G3, matching the references' intervals
and their 100ms spacing. The stream pair reuses those resolution notes as
single tones at ~55% the volume: they fire far more often during a broadcast,
so they have to sit under the call sounds rather than beside them.
Stream start is the one event that happens once per broadcast, so it can be a
chord — with a low thump and a short air layer that both die inside 150ms,
adding weight without length. Stream end mirrors it descending, quieter, and
without the air, since brightness reads as arrival.
Generators and the measured parameters are kept in tools/sfx so these can be
retuned without redoing the analysis. Total size drops from 150KB to 41KB.
Electron's audio: 'loopback' captures the whole output mix, this app's own
playback included — so everyone else's voices went back out inside the share
and each listener heard themselves. Not acoustic echo but a digital copy of
the output, which is why headphones never helped, and why shareAudio already
defaulted to off in the desktop app.
Electron offers no way to exclude our own audio: the docs allow only
'loopback' or 'loopbackWithMute', and the handler discards the renderer's
constraints (restrictOwnAudio never arrives). electron-native-screenshare does
it at the OS level — WASAPI process loopback on Windows — capturing only the
shared window when its pid resolves, and otherwise everything except us.
The module hands raw PCM to the main process, so it crosses IPC and is
scheduled onto a running cursor in Web Audio to become a MediaStreamTrack,
published as ScreenShareAudio. Loading is optional and failure degrades to a
silent share rather than blocking the app or the screen share.
The browser path is untouched: Chrome honours restrictOwnAudio and has no echo.
Verified by typecheck (web and Electron main) and the web suite. The audio path
itself cannot be exercised here — no Windows, no Electron, no audio device.
Soundboard: naming a clip used window.prompt, which Electron does not
implement — it returned nothing, the flow aborted in silence, and adding a
sound worked in the browser while doing nothing at all in the desktop app.
Replaced with a two-step field inside the popover, identical in both.
Spotify, three separate defects behind the two symptoms reported:
Out of sync — a 20s poll stacked on the activity store's 5s debounce left
everyone else on the previous track for up to 25s. The next poll is now
scheduled just past the current track's end instead of on a fixed interval,
and a track change bypasses the debounce (it happens once every few minutes;
the debounce exists for chatty producers).
Vanishing — a paused track, and the silent gap Spotify reports between two
songs, both cleared the activity outright. Pausing is now carried as state
rather than absence, and an empty answer is tolerated for 25s before the
block comes down.
Progress bar — timestamps are computed with the server's clock and were drawn
against the viewer's, so any drift displaced the bar; and it kept advancing
after a pause until the next poll. The ready payload now carries server time
so each client can correct its own offset, and the bar freezes when paused.
Tray, native notifications and system audio in screen share were all found
already implemented and wired end to end; recorded in the roadmap rather than
built again.
The add-sound control was gated on channelPermissions, which carries the
per-channel bitfield; MANAGE_SPACE lives in the space-level one. The check
silently evaluated false for everybody, including owners, so the button never
rendered and there was no way to add a sound at all.
Also raise the clip cap to 2 MB — a 30-second clip at a high bitrate cleared
1 MB — and stop reporting every upload failure as 'too large', which sent
people to shrink a file that was not the problem.
Soundboard: the trigger travels over the WebSocket and every client in the
call plays the clip locally, instead of mixing it into the presser's
microphone or publishing a LiveKit track. No upstream bandwidth, no media
stack changes, and the clip is not degraded by voice processing.
Fan-out uses a new sendToRoomParticipants rather than sendToRoom: the latter
broadcasts a space room to the whole space, which is right for the presence
the sidebar shows and wrong for anything audible. The cooldown is enforced
server-side — a client-side one only slows down people not trying to abuse it,
and a soundboard is the easiest thing here to turn into a weapon. Playing is
open to anyone in the call; deciding what the buttons are needs MANAGE_SPACE.
Account menu: the name in the user bar had cursor-pointer and no handler, so
the interface was already promising a click that did nothing. Offers profile,
status and copy-id — not the Clips or account switching the reference design
shows, which would be dead UI here.
Call timer: startedAt comes from the server, so a late joiner sees the call's
age rather than their own arrival. Empty space rooms are destroyed already,
which is what makes the next call start from zero — no reset logic needed.
Voice stays get their own table rather than joining the audit log: that table
records points in time, a call is an interval, and pairing join/leave point
events would leave every query guessing at joins whose leave never arrived.
Sessions are opened and closed inside joinRoom/leaveCurrentRoom rather than at
the seven call sites that reach them, so no path can be missed, and
destroyRoom closes them too — it bypasses leaveCurrentRoom and would otherwise
leak open rows.
A restart leaves sessions open with an unknowable end time. They are closed at
startedAt, discarding that time rather than inventing it: crediting the gap
would hand someone hours they never spent, and the numbers are the point.
Mirrors the existing users.status sweep on boot.
Only closed sessions count, so a figure does not move on every refresh. Bars
scale to the leader, not the total — with five people every share of a total
looks identical. Statistics are readable by any member, since they are the
group's own numbers; the audit log, which names who did what, stays on
MANAGE_SPACE.
Records who changed what, and is the mechanism statistics will read — one
event table rather than two logs that drift apart.
The table is deliberately generic (action + target + JSON metadata) so a new
action needs no migration. Writes never throw: a kick must not fail because
its log entry could not be written, since the kick already happened.
Leaving is recorded as a different action from being removed. The same route
serves both, and a log that conflates them misleads exactly when it matters.
Actor is nullable with ON DELETE SET NULL: the event outlives the account, and
a log that vanished with its actor would be worthless. Reads are gated on
MANAGE_SPACE rather than a new permission bit, which would default to nobody
until every role was re-edited. Paging uses the snowflake id, stable even for
two events in the same millisecond, and an action this build does not know
still renders a row.
Favourites are stored server-side per user, so one made on the phone is there
on the desktop — the point of favouriting. The whole result is stored rather
than an id: the provider offers no lookup by id, so an id-only favourite could
not be rendered without re-finding it through search.
Category chips translate their label but not their query, which goes to a
provider that indexes in English.
The star sits beside the tile button rather than inside it: a button within a
button is invalid and swallows the click. Toggling is optimistic and reverts
on failure, and favourites skip both the loading skeleton and the infinite
scroll, which belong to provider-backed browsing only.
Server caps favourites per user and rejects non-http(s) URLs, which become
<img src> in everyone's picker.
OAuth Authorization Code flow, with tokens kept server-side: refreshing needs
the client secret, so the browser never holds a Spotify token — it asks this
instance what is playing and this instance calls Spotify.
The callback arrives as a plain browser redirect with no Authorization header,
so the OAuth state carries the user id signed with the instance secret and is
compared in constant time; without that, anyone could bind their Spotify
account to another user.
Activities are now tracked per producer. pushActivities replaced the whole
list, so the desktop game detector and Spotify would erase each other — losing
exactly the case this is for, a game and Spotify at once.
Polling backs off when the tab is hidden and keeps the last known track on a
network error rather than reporting 'stopped listening'. A rejected refresh
token (access revoked on Spotify's side) drops the row so the UI stops
claiming a live connection.
Scope is read-only: user-read-currently-playing and user-read-playback-state.
Per the fork's language rule, the new UI ships in en and pt-BR, and this
round also translates the privacy panel.
Nothing in the project was translatable — every string sat inline in English.
en.ts is the source dictionary and its type is derived from it, so a typo or a
missing key fails typecheck instead of rendering the raw key at runtime.
pt-BR.ts is deliberately Partial: translation proceeds one system per update
and anything absent falls back to English, so a half-migrated interface is
never broken, only partly English.
Locale is persisted, guessed from the browser on first run, and kept in sync
with <html lang> through a subscription — persisted state rehydrates after
first paint, so a one-off assignment would miss it.
Translates the voice input panel (including the mic test shipped earlier
today) and the profile card as this round's system. Language options are
labelled in the active language, so a wrong pick can always be undone.
The activity selector sat below `if (!isOpen || !user) return null`. `user`
arrives asynchronously, so the hook ran on some renders and not others; React
counts hooks per render and tore the tree down with error #310 as soon as a
profile finished loading.
Move it above the guard and let the selector tolerate a null user. Typecheck
and the suite both passed with the bug in place — TypeScript cannot see hook
order and nothing renders this modal across the null-to-loaded transition.
The activity pipeline was already complete end to end — Activity type, store,
WS broadcast, server validation, presence relay, and an ActivityCard used by
four list surfaces — but the profile card rendered none of it, which is the
'Listening to Spotify' block the design calls for.
Add ProfileActivity: richer than ActivityCard because the card has room for
artwork, track and artist, so it reads details/state/assets. All optional, so
it degrades to the bare name that today's process-based detector supplies.
Also scheme-check activity image assets server-side. activity.url was already
restricted to http(s) but assets.largeImage/smallImage were only length-checked
— an asymmetry that was harmless while nothing rendered them, and is not once
they become <img src>: a client could point them at a host it controls and
harvest the IP of everyone opening that profile.
The level meter only measured a stream a call had already opened, so settings
offered no way to check a mic before joining — the panel said as much.
Add startMicTest/stopMicTest on AudioManager: the processed input bus is
routed to the master output through a dedicated gain node, so the loopback can
be disconnected precisely. Settings had deliberately never opened the mic
itself; a mic test cannot honour that, so the test hands the mic back when it
stops.
Releasing needs two independent guards, because the user may join a call
mid-test: AudioManager only stops the exact stream it opened (identity check,
not a flag), and the caller must consent — the UI reads the call state, which
AudioManager cannot, as it does not import stores. Unmounting mid-test tears
the loopback down too.
The profile popout already existed and was reachable from eleven places —
messages, mentions, avatars, member list, DMs, activity panel — but no voice
surface opened it, so clicking someone during a call did nothing.
Wire it into the voice user rows (VoiceChannel's sidebar list) and the name
label on grid tiles, whose avatar was already a ProfileAvatar; the name beside
it not reacting read as the click failing.
Left mobile alone deliberately: MobileSpacesScreen already opens the profile
from its row wrapper, and MobileVoiceJoinSheet would layer a history-pushed
full-screen profile inside a bottom sheet, which cannot be verified here.
The composer's GIF button drew a filled rounded rect with the letters knocked
out, which reads as a solid square rather than a picker. Invert it: stroked
outline with filled letters, reusing the original glyph paths scaled to centre.
Banners already accept absolute URLs on both ends (server isValidAssetUrl
allows http(s); the profile render branches on banner.startsWith('http')), so
the picker stores the remote URL directly with no upload path. Previews can now
hold either a blob: or an https: URL, so revoking is guarded — calling
revokeObjectURL on a remote URL is a silent no-op that would hide a mistake.
The channel name under 'Voice Connected' was a plain div. Making it navigate
needed more than an onClick: voiceStore never recorded which space the call
was in, and spaceStore.channels only holds the space currently being viewed —
so after navigating away the call's channel was unresolvable, which is also
why the label degraded to a generic 'Voice Channel'.
Capture space and channel name at join time (the only moment they are
reliable) and use them for both the label and the jump. Covers space calls
and DM calls.
Avatar opened the profile popout whenever it received a user prop. Since user is how every avatar gets its gradient, colour and status dot, all 22 call sites became profile triggers by accident — including the picture inside the profile card itself, which re-anchored the card to that picture on every click and walked it across the screen (120px right, 36px down, until it pinned at the viewport clamp).
Avatar is now presentational. A new ProfileAvatar carries the open-the-profile behaviour at the five call sites that actually want it. The card's own picture escalates to the full profile modal instead of reopening the card.
The card also places itself off its measured size via the shared computeFloatingPosition engine, replacing six call sites that each hand-computed coordinates against a guessed 460px card height.
Closes#37
* ci: run typecheck, build & tests on PRs and main
Add a GitHub Actions workflow (.github/workflows/ci.yml) that installs on
Node 20 (the pinned runtime), builds shared/server/web, typechecks desktop,
and runs the full vitest suite across all packages on every pull request and
push to main. The repo already had 500+ tests but only ever ran them locally.
Also wires up the missing test/typecheck plumbing this exposed:
- web package had 55 test files but no test script — add test/test:watch/typecheck
- add root-level test and typecheck aggregate scripts
- fix a latent web type error the new typecheck surfaced: reference
vite-plugin-pwa/react types so virtual:pwa-register/react is typed
(useRegisterSW callback params were implicit any; the tsc half of the web
build script has been failing, masked by vite build ignoring type errors)
* test: fix two latent failures surfaced by CI on Node 20
- server: exclude dist/** from vitest. Vitest 4's default exclude dropped
dist/, so after `pnpm build` emits compiled .test.js files, vitest ran those
stale copies alongside src/*.test.ts and they failed (compiled vi.mock paths
resolve differently).
- web: polyfill Blob.prototype.stream in the jsdom test setup. jsdom's Blob has
no .stream() on Node 20 (Node 25 provided one, masking this locally); undici's
Response constructor calls blob.stream(), so new Response(blob) threw
'object.stream is not a function'.
computePermissions() returned the space @everyone role's permissions without
verifying the caller had joined the space. Because CREATE_INVITE is in
DEFAULT_EVERYONE_PERMISSIONS, any authenticated user could mint an invite code
for a request-only space — whose id is listed by /api/spaces/explore — and then
self-join via /api/spaces/:id/join, bypassing the join-request approval flow.
The same gap let non-members read message history and search default channels.
Root cause:
- computePermissions now returns 0n for non-members (space owner and instance
admin still short-circuit first, so they are unaffected).
Defense in depth (request-only spaces are approval-gated, never invite-joinable):
- both invite-code join endpoints reject visibility='request' (private stays
invite-joinable — its only entry path; public too).
- POST /api/spaces/:id/invite refuses to hand out a code for request spaces.
- POST /api/dm/space-invite refuses to card a local request space, checked by
space id against the local table so a spoofed spaceInstanceOrigin can't slip
past it.
- InviteModal hides the invite affordances for request spaces.
Also removes the unused computeCategoryPermissions(), which duplicated the
resolution algorithm without the membership gate.
Adds unit + route + component tests covering non-member/member/owner/admin
resolution and the request/private/public visibility matrix.
Reported-by: BadAtCaptchas (#2)
Finding 1: handleReactionAdd/Remove now drop reactions on a dead 1-on-1
(isDeadOneOnOne) — previously a survivor could react on a Deleted-User
thread and the relay fanned out to all peers via undefined target-origins.
Client Message.tsx withdraws add/toggle reaction affordances for dead DMs
(existing reactions still display read-only).
Finding 2: dmMembership purge test now runs with foreign_keys=ON (matches
prod) and asserts dm_members/dm_messages cascade cleanup on channel purge.
Finding 3: tombstone group-DM ownership transfer filters isDeleted=0 so
ownership can never move to a tombstoned member; covered by a new test.
devsyncwrldandClaude Opus 5