Hook POST /api/dm/:id/members and DELETE /api/dm/:id/members to the
federation outbox so membership changes propagate to peer instances:
- Assign federatedId when a group DM first gains remote members
- Relay member_add events with group metadata and participant-aware routing
- Relay member_remove (leave) events to peer instances
- Relay ownership_transfer when the departing user was the group owner,
updating the federated owner columns accordingly
Uses messageHomeInstance to determine whether the target message
originated locally (lookup by ID) or was replicated from another
instance (lookup by sourceInstance + sourceMessageId). Supports
N-instance federation and falls back to relay-sender lookup for
backward compatibility with in-flight outbox entries.
When Instance 2 receives a relay from Instance 1, it no longer
broadcasts dm_message_created to members whose home instance is Instance
1. Those users already have the original message via their home WS
connection. This prevents duplicate messages from appearing in the
sender's chat.
Reaction outbox events used reactionId (add) or a composite dedup key
(remove) as the event messageId. The receiver looked up dm_messages by
sourceMessageId = event.messageId, which never matched because it was
searching for a reactionId, not the actual message ID.
Fix: include the actual DM messageId in the reaction payload JSON.
The receiver now uses event.reaction.messageId for the lookup, with
fallback to event.messageId for backward compatibility.
The source instance's thumbnail file doesn't exist locally, so copying
its thumbnailFilename causes a 404 in the chat view. Set to null so
AttachmentRenderer falls back to the full image URL.
Insert attachment rows with sourceUrl as interim filename so the initial
WebSocket broadcast includes working remote URLs. AttachmentRenderer
already handles http:// filenames as direct URLs. The file worker will
UPDATE the filename to the local path after download completes.
This fixes the invisible bubble bug where attachment-only messages
(content: null, attachments: []) appeared as empty messages.
The relay was failing because processCreateEvent relied on the friends
table to discover the DM recipient, but friendships aren't federated
across instances. Also, resolveLocalUser matched deleted replicated
users before active ones.
- Add participants[] to FederationRelayEvent with homeUserId/homeInstance
for all DM channel members
- Add getDmParticipants() helper to look up member identities
- Include participants in outbox payloads (create/update) and sync events
- Rewrite processCreateEvent to resolve participants directly, compute
canonicalDmPairId, and findOrCreateDmChannel — removing the entire
friends-list fallback (60+ lines)
- Fix resolveLocalUser to filter out deleted users (is_deleted = 0)
and prefer the replicated user match when multiple candidates exist
The old query missed channels where both members appear as 'local'
(e.g., native Nova + native youruser on orbit). Using
canonical_pair_id is simpler and correct — any channel with a pair
ID is a federated DM that should be synced.
Messages by replicated users (e.g., Jannis browsing orbit) have
source_instance IS NULL because they were created on that instance.
The sync endpoint was incorrectly filtering these out, causing most
DM history to be missing from the sync response.
Task 13: Hook server-to-server peering into connectToRemote (non-fatal)
and add federation API namespace to the client (initiate, peers, revoke).
Task 14: Show a transparency notice in the DM welcome header when the
other user is on a remote instance, informing that messages are stored
on both home instances and are not end-to-end encrypted.
Task 15: Add Federation section to the instance settings General panel
with DM relay toggle, TTL config, and a live peer list with revoke
buttons. Also extends InstanceAdminSettings type and the server settings
route to support federationRelayEnabled / federationRelayTtlDays.
Add POST /api/federation/sync for peer instances to catch up on missed
DM mutations after downtime. Uses HMAC-SHA256 authentication (same as
relay), determines shared channels between peers via raw SQL join on
dm_members/users, queries the federation_mutation_log with timestamp
pagination, and returns FederationRelayEvent payloads for each mutation
type (create, update, delete, reaction_add, reaction_remove).
Add federation routes module with four endpoints:
- POST /api/federation/peer/initiate — admin-only, starts peering handshake
with a remote instance (validates origin, generates HMAC secret, sends
server-to-server accept request with 10s timeout)
- POST /api/federation/peer/accept — server-to-server first contact,
rate-limited to 10 req/min per IP, creates/activates peer record
- GET /api/federation/peers — admin-only, lists all peers (hmacSecret excluded)
- DELETE /api/federation/peers/:id — admin-only, revokes peer and cleans outbox
Lower minimum query length from 2 to 1 character so single-letter
searches return results.
Fix dedup to prefer native profiles (homeUserId=null) over replicated
ones. Previously the first-seen result won, which was usually the
local replicated profile (no instance badge, namespaced username).
Now when a native profile is found on the remote instance, it replaces
the replicated copy — showing the clean username with the instance badge.
The 500 MB cap was arbitrary. Self-hosted admins should decide based
on their own disk space. Global multipart ceiling raised to 5 GB,
validation cap to 5120 MB.
Reads max_upload_size_bytes from instance_settings per-request and
passes it to request.file() so Fastify kills the stream at the
admin's configured limit. Checks file.truncated to properly reject
files that exceed the limit instead of saving corrupted data.
Adds cleanupOldMedia() to the storage janitor and a new endpoint
POST /api/admin/storage/cleanup-media { maxAgeDays, dryRun }.
Deletes chat attachments older than the specified threshold while
preserving profile images.
When a federated user's local password hash is stale (e.g. they changed
their password on the home instance and sync failed), the login handler
now falls back to verifying credentials against the home instance. If the
home instance accepts the password, the local hash is silently updated
without touching passwordChangedAt, so existing valid JWTs remain valid.
Add homeInstance, role, joinedAfter, joinedBefore, and sort query
params to GET /api/admin/users. Add GET /api/admin/users/instances
endpoint returning distinct home instance domains.
- Import BITRATE_MATRIX_KBPS from shared constants
- Parse and return bitrateMatrixOverrides in rowToLimits
- Widen updateData type to accept null values
- Raise maxBitrateKbps and minBitrateKbps caps from 50 000 to 1 000 000 kbps
- Add full validation for bitrateMatrixOverrides (null to clear, key/value checks)
Replace hardcoded VALID_RESOLUTIONS/VALID_FRAMERATES with shared STANDARD_RESOLUTIONS/STANDARD_FRAMERATES constants. Update rowToLimits to handle the 'native' resolution value in DB parsing, and update all PATCH validation blocks to accept the expanded resolution/framerate sets and the 'native' option.
- Add GET/PUT/DELETE /api/categories/:id/overrides endpoints with
privilege escalation guard matching existing channel override pattern
- Add isCategoryPrivate() and broadcastCategoryOverrideChange() helpers
that cascade visibility changes to all channels in a category
- Fix PATCH /api/channels/:id to recompute permissions via
broadcastOverrideChange when categoryId changes (different category
overrides may apply)
- Include isPrivate flag on categories in GET /api/spaces/:id payload
using batch-fetched @everyone overrides
- Include isPrivate on categories in broadcastChannelLayout and
PATCH /api/categories/:id broadcasts
- Accept showActivity boolean in PATCH /api/users/@me body
- Validate and persist as integer in DB
- Update ConnectionManager cache on toggle
- Broadcast empty activities via presence_update when toggled off
- Add Activity, ActivityType, ActivityTimestamps, ActivityAssets types to shared types
- Add activity_update client event and activities field on presence_update server event
- Add userActivities to ready payload and showActivity to User/UpdateUserRequest
- Create shared activities.ts with ACTIVITY_LIMITS, ACTIVITY_PRIORITY, getPrimaryActivity
- Add show_activity column to users table (schema + migration)
- Update sanitizeUser with isSelf parameter; only include showActivity for self
- Fix .map(sanitizeUser) calls to use arrow wrapper to prevent index-as-boolean bug
- Mark auth routes (register/login) as isSelf=true since they return own user data
1. Video aspect ratio: remove container border/overflow-hidden, use
preload="metadata" so browser knows dimensions before play
2. YouTube Error 153: remove sandbox attr (too restrictive), add full
allow permissions (encrypted-media, accelerometer, gyroscope, etc.)
3. Google Images not displaying: detect image Content-Type from HTTP
response in metadataFetcher, override classifier to create image
embed for URLs that serve image/* content
4. Audio seeking broken: add HTTP Range request support in uploads
route (Accept-Ranges, Content-Range, 206 Partial Content)
- Fix ?? to || in metadataFetcher.ts to handle empty strings from Cheerio
- Fix stale embeds on message edit: delete old embeds synchronously before
broadcast, then resolve new ones async (all 4 edit paths: REST+WS, msg+DM)
- Revert unrelated MessageList.tsx scroll threshold change (5000 not 150)
- Remove duplicate embed indexes from migrateAddIndexes (kept standalone ones)
Integrate embed infrastructure into the complete message flow:
- messages.ts: batch-fetch embeds in GET, resolve on POST, re-resolve on PATCH
- dm.ts: same pattern for DM messages with isDm=true
- search.ts: include embeds in all 4 search/around endpoints
- events.ts: embed resolution in WS message create/edit for both space and DM
- Fix embedClassifier.ts type errors (regex match undefined → null)
- Add embeds: [] to all inline MessageWithUser/DmMessageWithUser constructions
- Add embeds: [] to chatStore optimistic message
profileSync stored avatar/banner paths with /api/uploads/ prefix on remote
instances, causing resolveAssetUrl to produce double-path URLs like
https://remote/api/uploads//api/uploads/file.jpg that 404'd. Store bare
filenames instead, strip prefix defensively in resolveAssetUrl and server-side
for existing data self-healing.
Root cause: own messages echoed by the server marked channels unread when
the user had already navigated away. Seven related bugs compounded the
problem — stale read states, missing cleanup on space/DM removal, REST
broadcast ignoring VIEW_CHANNEL, and no validation on channel_ack writes.
Frontend:
- Skip markChannelUnread for the user's own messages (federation-aware)
- Walk backward past temp_ IDs in ackChannel instead of bailing
- Re-fire ack timer when temp message is replaced by server-confirmed ID
- Add removeChannelStates to clean up unread/read/message caches
- Clean up chatStore on removeSpace, removeDmChannel, removeInstanceSpaces
Server:
- Use sendToChannel instead of sendToSpace for REST message creation
- Clean up read_states on space deletion, member kick/leave, and ban
- Validate channel membership before accepting channel_ack writes
- Clean up read_states on DM leave and DM channel deletion
- Resize avatars/icons to 256px and banners to 1280px (client crop + server safety net)
- Add server-side resizeProfileImage() for federation/API uploads without crop modal
- Fix unconstrained crop on RegisterPage and CreateSpace (was missing maxOutputDimension)
- PWA: switch to autoUpdate with skipWaiting/clientsClaim for seamless deploys
- Storage janitor: exclude profile images from unlinked cleanup, delete stale thumbnails
- Add deleteAttachmentByFilename() to clean orphaned attachment records for profile images
- Migration: one-time cleanup of stale profile image attachment records
- GeneralPanel: wrap in <form> to prevent implicit submission
isValidAssetUrl() was rejecting bare filenames (e.g. "1234567890.webp") which
is the established convention the frontend sends. Now accepts bare filenames
while still blocking path traversal and unsafe schemes.
Also updates client-side password validation to match server's 8-char minimum.
- SSRF protection: DNS resolution + private IP blocking on metadata fetcher
- Upload security: CSP/X-Frame-Options headers, SVG forced download, nosniff
- Auth hardening: JWT secret min length, password min 8 chars, token revocation via password_changed_at
- Attachment ownership verification before linking to messages
- Message length limit (4000 chars) enforced on client and server
- Asset URL validation on avatar/banner updates
- Federation instance validation (domain regex, origin scheme, length limits)
- DB indexes on all FK columns for query performance
- Migrations: nullable moderator columns, dm_messages reply_to FK constraint
- File cleanup on avatar/banner replacement and space deletion
- Fastify trustProxy, AbortController on fetches, typing map size cap
Jannis Braun
