spreed/capabilities.md в 1e3122db66f96ca8039a3225f24b5e3765c6e0cb

title
Capabilities

3.0 (Initial Talk release)

guest-signaling - Guests can do signaling via api endpoints
empty-group-room - Group conversations can be created without inviting a Nextcloud user group by default

guest-display-names - Display names of guests are stored in the database, can be set via API (not WebRTC only) and are used on returned comments/participants/etc.
multi-room-users - Users can be in multiple conversations at the same time now, therefor signaling now also requires the conversation token on the URL.
chat-v2 - Chat messages are now Rich Object Strings and pagination is available, the previous chat is not available anymore.

favorites - Conversations can be marked as favorites which will pin them to the top of the conversation list.
last-room-activity - Conversations have the lastActivity attribute and should be sorted by that instead of the last ping of the user.
no-ping - The ping endpoint has been removed. Ping is updated with a call to fetch the signaling or chat messages instead.
system-messages - Chat messages have a systemMessage attribute and can be generated by the system
mention-flag - The conversation list populates the boolean unreadMention when the user was mentioned since their last visit
in-call-flags - A new flag participantFlags has been introduced and is replacing the participantInCall boolean.

invite-by-mail - Replaced by invite-groups-and-mails Guests can be invited with their email address
notification-levels - Users can select when they want to be notified in conversations
invite-groups-and-mails - Groups can be added to existing conversations via the add participant endpoint

locked-one-to-one-rooms - One-to-one conversations are now locked to the users. Neither guests nor other participants can be added, so the options to do that should be hidden as well. Also a user can only leave a one-to-one conversation (not delete). It will be deleted when the other participant left too. If the other participant posts a new chat message or starts a call, the left-participant will be re-added.
read-only-rooms - Conversations can be in read-only mode which means people can not do calls or write chat messages.

chat-read-marker - The chat can be optionally marked read by clients manually, independent from the loading of the chat messages.
webinary-lobby - See Webinary management for technical details.
start-call-flag - Only moderators or users might be able to start calls.

chat-replies - Normal chat messages can now be replied to. Check the isReplyable parameter on the message object.
circles-support - Conversations can be created for circles and all circle members can be added to existing conversations

config => attachments => allowed - Whether the user can upload files into a chat
config => attachments => folder - User defined folder where items should be uploaded to
config => conversations => can-create - Whether the user can create public and group conversations, if not only one-to-one conversations are allowed
force-mute - "forceMute" signaling messages can be sent to mute other participants.
conversation-v2 - ~~The conversations API v2 is less load heavy and should be used by clients when available. Check the difference in the Conversation API documentation.~~ Replaced by API v4 when conversation-v4 is set.
chat-reference-id - an optional referenceId can be sent with a chat message to be able to identify it in parallel get requests to earlier fade out a temporary message

conversation-v3 - ~~Whether conversations API v3 is available~~ Replaced by API v4 when conversation-v4 is set.
sip-support - Whether SIP can be configured and enabled by moderators. The conversations API will come with some new values sipEnabled which signals whether this conversation has SIP configured as well as canEnableSIP to see if a user can enable it. When it is enabled attendeePin will contain the unique dial-in code for this user.

chat-read-status - On conversation API v3 and the chat API the last common read message is exposed which can be used to update the "read status" flag of own chat messages. The info should be shown only when the user also shares their read status. The user's value can be found in config => chat => read-privacy.
listable-rooms - Conversations can searched for even when not joined. A "listable" attribute set on rooms defines the scope of who can find it.
phonebook-search - Is present when the server has the endpoint to search for phone numbers to find matches in the accounts list
raise-hand - Participants can raise or lower hand, the state change is sent through signaling messages.
room-description - A description can be get and set for conversations.
config => chat => read-privacy - See chat-read-status
config => previews => max-gif-size - Maximum size in bytes below which a GIF can be embedded directly in the page at render time. Bigger files will be rendered statically using the preview endpoint instead. Can be set with occ config:app:set spreed max-gif-size --value=X where X is the new value in bytes. Defaults to 3 MB.

delete-messages - Allows to delete chat messages up to 6 hours for your own messages or when being a moderator. On deleting the message text will be replaced and a follow up system message will make sure clients and users update it in their cache and storage.
rich-object-sharing - Rich objects can be shared to chats. See OCP\RichObjectStrings\Definitions for more details on supported rich objects and required data.
conversation-call-flags - Whether the room api provides the callFlag to tell apart video and audio calls

conversation-v4 - Whether conversations API v4 is available. This also means that v1, v2 and v3 are not available anymore. The changes on API v4 allow a user to have multiple sessions in the same conversation on different devices which is incompatible with the data structure that was used by the previous APIs.