Adds a `type` column (TEXT NOT NULL DEFAULT 'user') to the dm_messages
table via schema, migration, and type definition. Updates
buildDmMessageWithUser and the inline replyTo builder in the GET
messages handler to include the field in all DM message responses.
- Use inline badges for file/audio attachments (next to file size)
instead of absolute positioning which overflowed the container
- Keep overlay badges for images/video (overflow-hidden containers)
- Match tooltip text to toast notification text for consistency
- Fix critical: outbox worker now copies file_rejected payload fields
(attachmentId, sourceFilename, rejectionReason, rejectionLimit,
affectedUserIds) so the reverse relay actually delivers them
- Fix: add sourceFilename to file_rejected event for reliable
multi-attachment matching on the sender side
- Fix: change text-accent-warning to text-accent-amber (valid class)
- Add Array.isArray guard on federationMeta parse
Both buildDmMessageWithUser and buildMessageWithUser manually map
attachment fields — add federationStatus and federationMeta so they
reach the frontend.
The leave-group-DM ConfirmDialog in the space/channel sidebar return
path was unreachable — group DMs only render in the DM view branch,
so setLeaveGroupDmId is never called from the space view.
The X button on group DMs now opens a ConfirmDialog instead of
soft-closing. The right-click 'Leave Group' context menu also
routes through the same confirmation. Prevents dead group DM data
from accumulating when users soft-close instead of leaving.
members.length > 2 caused group DMs to lose their identity when a
member left and only 2 remained. ownerId is the canonical marker set
at group creation and persists through the group's lifecycle.
The buildReadyPayload query fetched all dm_members rows without
checking the closed flag, causing closed DMs to reappear on every
page reload. The REST endpoint GET /api/dm already filtered correctly.
Federated 1-on-1 DMs showed the raw snowflake ID as the display name
and no avatar when the remote user had no pre-existing local record.
processCreateEvent used resolveLocalUser (find-only) instead of
resolveOrCreateReplicatedUser, and relay events carried no profile data
for participants.
- Add profile snapshot (displayName, avatar, avatarColor) to
FederationRelayParticipant and populate it in getDmParticipants
- Change processCreateEvent to auto-create replicated user stubs and
hydrate them with profile data from the relay event
- Fix hydrateReplicatedUserProfile URL resolution for homeInstance
values without protocol prefix
- Fix WelcomeHeader: return null while DM data is loading (eliminates
"unknown" flash on reload), use displayName for @mention text
Federated relay messages can have local snowflake IDs that don't match
chronological (createdAt) order — a message sent earlier on a remote
instance can arrive later and get a higher local ID. This caused a
permanent mismatch between the ready event's lastMessage (MAX id) and
the acked message (last in createdAt display order), making federated
DM channels appear unread after every server restart.
- Server: change ready event DM lastMessage query from MAX(id) to
ORDER BY created_at DESC (matching the DM REST API)
- Frontend: change ackChannel to ack MAX(id) among loaded messages
instead of last in display order (consistent with server comparison)
- Fix federated username display fallback in UserDiscoverCard
When converting a 1-on-1 DM to a group, the existing DM partner was
incorrectly required to be your friend. DMs don't require friendship,
so this check was over-strict. Added fromDmChannelId parameter to
createGroup — existing members of the source DM are exempt from the
friendship validation.
Replace global user search with socialStore friends list, fixing
federation ID mismatch that caused false 'not your friend' errors.
Add multi-select with chips, immediate friend list display, and
client-side search filtering.
Use parseFederatedUsername() consistently to show base username instead
of raw user@domain format. Replace inline domain text in Username
component with compact globe icon + tooltip. Add globe icons to DM
header and sidebar for federated users.
Bare filenames stored on replicated user stubs can't be resolved by
the home WS (normalizeUserAssets only runs for remote origins).
Now resolves avatar/banner to absolute URLs pointing to the user's
home instance so they render correctly without page refresh.
Also overwrites stale bare filenames from the prior deploy.
Replicated user stubs created by resolveOrCreateReplicatedUser had
null avatar/displayName, causing blank profiles in the UI until
page refresh. Friend relay events now carry profile snapshots
(displayName, avatar, avatarColor, banner, bio) so the receiving
instance can hydrate stubs with real data.
The friend_request_accepted and friend_removed WS handlers updated
socialStore but not discoverStore, so the discovery page buttons
stayed stale until page refresh. Decline/cancel already updated
both stores correctly.
Rename appendMutationLog and queueOutboxEvent params from DM-specific names
(dmMessageId/dmChannelId/messageId) to generic (entityId/contextId) with a new
contextType param defaulting to 'dm'. Update all internal schema column references
to match the renamed outbox/mutation-log schema columns. Add buildFriendContextId
and getFriendEventTargets helper functions for friend event relay routing.
Renames DM-specific columns in federation_outbox (dm_channel_id → context_id, message_id → entity_id) and federation_mutation_log (dm_message_id → entity_id, dm_channel_id → context_id) to generic names, adding context_type = 'dm' for all existing rows so the outbox can carry friend events too.
Adds migrateResetFederationSyncForLegacyDms which resets last_synced_at=0
on all active federation peers so the S2S sync worker re-pulls all mutation
log entries (including newly-backfilled legacy DMs) on next server startup.
A legacy_dm_sync_done flag on instance_settings ensures this runs exactly once.
Replace myUserId with full myUser object so isSelf() can handle cross-instance
identity matching when finding the other 1-on-1 DM member. Pass GroupDmUserIdentity
objects (id, homeUserId, homeInstance) to api.dm.createGroup instead of raw id strings.
Rewrite the group DM creation endpoint to accept identity objects
(GroupDmUserIdentity) instead of raw user ID strings. Each identity
is resolved to a local database user via resolveOrCreateReplicatedUser
for federated users or direct ID lookup with resolveLocalUser fallback
for local users. Dedup and caller-exclusion checks now operate on
resolved local IDs rather than input IDs.
Make both identity resolution helpers module-level exports so the group DM
endpoint can import and use them when resolving federated user identities
during group DM creation.
Replace flat userIds array with a structured users array of GroupDmUserIdentity
objects, enabling the group DM endpoint to accept federated identity hints
(homeUserId, homeInstance) alongside the local user ID.
addDmChannel must be called before navigate() so the route effect
finds the channel in the store immediately. Without this, the chat
view renders empty until the WS dm_channel_created event arrives.
Also show an error instead of silently returning when otherMember
cannot be determined.
- Add CreateGroupDmRequest to API client imports, type declaration, and implementation (POST /dm/group)
- Update AddDmMemberModal to branch on channel type: 1-on-1 DMs call createGroup and navigate to new channel, group DMs call addMember directly
- Remove legacy getApiForOrigin/channelOriginMap usage — S2S federation relay is now entirely server-side
- Hoist callerUser DB query above the federation block so it's fetched
once and reused for response building, federation ID assignment, and
relay payload construction (was fetched 3 times).
- Add homeInstance !== domainOrigin guard to finalTargets augmentation,
matching the existing pattern in POST /api/dm/:id/members.
Adds a dedicated endpoint for creating group DMs with 3-10 members.
Validates friendship, deduplication, and member caps. Includes federation
relay support for remote instance members.
- Add migrateFixOneOnOneOwnerIds migration to NULL-out ownerId on all
existing 1-on-1 DMs (those with exactly 2 members)
- Fix POST /api/dm to create 1-on-1 channels with ownerId=null instead
of the creator's ID
- Guard POST /api/dm/:id/members: reject with 400 if channel has no
owner (i.e. is a 1-on-1), directing callers to POST /api/dm/group
- Guard DELETE /api/dm/:id/members: replace member-count check with
ownerId check; remove now-duplicate dmChannel query in that handler
- Add CreateGroupDmRequest type to shared types
The CREATE INDEX on federated_id in createTables runs before migrations.
On existing installs the column is still canonical_pair_id, causing a crash.
Wrap in try-catch so the migration handles it instead.
Replace the hard-delete in the leave handler with a soft-delete (sets
deleted_at timestamp) when the last member leaves a group DM. A new
janitor sweep in the federation worker runs hourly and purges channels
whose grace period has expired, cascading through reactions, embeds,
attachments, messages, members, outbox/mutation-log/file-queue entries,
and finally the channel itself.
All client-facing dm_channels queries now filter on deleted_at IS NULL
to hide soft-deleted channels from the REST API and WebSocket ready
payload.
In processCreateEvent, branch on event.federatedId: group DM messages now
look up the pre-bootstrapped local channel by federatedId instead of
computing a pair hash from two participants. In queueDmRelay, fetch the
channel's federatedId and ownerId and include federatedId in the outgoing
relay payload for group DMs so receiving instances can route correctly.
Add three new event processors to the relay endpoint for handling
inbound group DM membership changes from peer instances:
- processMemberAddEvent: handles member additions with channel bootstrap
(creates the local channel from group metadata on first contact)
- processMemberRemoveEvent: handles member removals with read state
cleanup and soft-delete GC when no local members remain
- processOwnershipTransferEvent: updates local channel owner tracking
All three are wired into the relay endpoint's event dispatch switch.
getGroupDmTargetOrigins() was called after the leaving user's dm_members row
was deleted, causing their instance to be excluded from member_remove and
ownership_transfer relay fan-out if they were the only member from that peer.
Move the target origins and leaving user lookups to before the deletion.
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
Extend FederationRelayEvent with member_add, member_remove, and
ownership_transfer event types, plus federatedId and the new optional
membership/ownership/group payload fields. Add FederationMembershipPayload,
FederationOwnershipPayload, and FederationGroupPayload interfaces.
Add federatedId to FederationSyncRequest for group DM sync filtering.
Remove federated_id from addMissingColumns so it no longer races with
migrateDmChannelsFederatedId on installs that have canonical_pair_id.
Wrap the table rebuild in PRAGMA foreign_keys = OFF/ON with a transaction
so the DROP TABLE succeeds even when dm_members/dm_messages hold FK refs.
Renames dm_channels.canonical_pair_id → federated_id and adds
owner_home_user_id, owner_home_instance, deleted_at columns to support
federated group DMs. Migration handles upgrade from existing installs
(full table rebuild to rename), partial states (ALTER TABLE), and fresh
installs (correct DDL from the start). Backfills federated_id for
existing 1-on-1 DM channels and owner federation identity for group DMs.
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.
Resolves canonical messageId (sourceMessageId for replicated messages,
local ID for originated messages) and includes messageHomeInstance so
the receiver can deterministically find the target message regardless
of which instance created it.
Consolidates 4 inline constructions of the instance origin URL into a
single shared function. Removes the private copy in federationWorker
and two ad-hoc domainOrigin variables in federationOutbox.
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.
When a user has federated WS connections, they can receive both the
original message (from their home instance) and the relay copy (from the
remote instance). The relay copy has a different message ID but carries
sourceMessageId pointing to the original. The dedup now checks both
directions: incoming relay copy vs existing original, and incoming
original vs existing relay copy.
SQLite stores self-referencing FK identifiers with quotes (REFERENCES
"dm_messages"), but the migration guard only checked for the unquoted
string. This caused the migration to re-run on every server startup,
which DROP TABLE dm_messages and triggered ON DELETE CASCADE on
attachments.dm_message_id — wiping every DM-linked attachment.
When a relay-created dm_message_created arrives for an unknown channel,
check if an existing DM has the same member (by homeUserId). If so,
route the message to the existing channel instead of adding a duplicate
sidebar entry. Prevents the transient duplicate that required page
refresh to dedup.
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.
After the file worker downloads a federated attachment, generate a
thumbnail using the same sharp pipeline as local uploads. This ensures
federated images use bandwidth-efficient thumbnails in chat view
instead of serving the full-size image.
Also fixes processCreateEvent to not copy the source instance's
thumbnailFilename (which doesn't exist locally).
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.
The file worker now UPDATEs the attachment row created by processCreateEvent
(swapping sourceUrl interim filename to local path) instead of INSERTing a
duplicate. Falls back to INSERT for legacy queue entries.
After download, broadcasts dm_message_updated so clients see the attachment
transition from remote hotlink to local file in real-time.
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.
Single function builds complete relay payload with attachments (including
sourceUrl) and participants. Eliminates 4-way duplication that caused
the attachments: [] hardcode bug.
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.
- buildRelayPayload now uses config.domain for local users instead of
empty string, so the relay receiver can resolve the user
- Fixed canonical_pair_id backfill: SQLite NULL || ':' || x = NULL,
so use COALESCE(home_user_id, id) instead of concatenation
Existing 1-on-1 DM channels created before federation had canonical_pair_id = NULL,
causing the relay to create duplicate channels on receipt of a message. On server
startup, migrations now: (1) compute and set canonical_pair_id for all existing
1-on-1 channels using home_user_id (falling back to regular id for local users),
and (2) merge any duplicate channels that share the same canonical_pair_id by
moving all messages, read states, and members into the oldest channel and deleting
the stale duplicate. Both steps are non-fatal and idempotent.
Federation relay was incorrectly defaulting to disabled, requiring
admins to discover a buried settings toggle. Since federation itself
is opt-in (you connect instances manually), relay should be on by
default. Also migrates existing instances from 0 → 1.
Adds a post-deploy step that removes Docker images and build cache
older than 24 hours. Prevents the gradual disk bloat from repeated
docker compose --build invocations (was 71GB on the Pi).
Jannis Braun