Skip to navigation

Release Notes

Version 7.1.0 - April 07, 2021

Changes

  • rtcstats:report is deprecated. Please use rtcstats:analyticsinstead
application.on('rtcstats:analytics', ({
  type: 'mos' || 'mos_report',
  mos: string,
  rtc_id: string,
  mos_report?: {
    average: string,
    last: string,
    max: string,
    min: string
  },
  report?: RTCStatsReport
}) => {}
  • Add logs reporter for remote logging
  • Add custom getStats parser
  • Remove callstats library and implementation

Version 7.0.2 - March 30, 2021

Fixes

  • Emit 'call:status:update' event when application offline and call disconnected

Version 7.0.1 - February 04, 2021

Fixes

  • Remove unnecessary warning message for call status transitions

Version 7.0.0 - February 02, 2021

Breaking changes

  • legs endpoint should be included in acl paths on JWT token creation
"acl": {
  "paths": {
    ...
    "/*/legs/**": {}
  }
}

Changes

  • Improve callServer setup time by pre-warming leg
  • Add the rtcObject and remote stream to the NxmCall object

Fixes

  • Return ClientDisconnected reason when client logouts from SDK

Version 6.2.1 - December 24, 2020

Fixes

  • Revert back to 6.1.1 from 6.2.0-alpha

Version 6.1.2 - December 24, 2020

Fixes

  • Update Typescript definitions

Version 6.1.1 - December 09, 2020

Fixes

  • Update Typescript definitions

Version 6.1.0 - December 01, 2020

Changes

  • Add enableEventsQueue flag to client configuration (default to true)
  • Add internal eventsQueue mechanism to guarantee order of events received during a session

Version 6.0.19 - November 26, 2020

Changes

  • Update reconnectionDelay to 2000 in socket.io
  • Add randomizationFactor to 0.55 in socket.io

Version 6.0.18 - November 19, 2020

Changes

  • Update documentation

Fixes

  • Add reconnectionDelay to 3000 and reconnectionDelayMax to 15000 in socket.io

Version 6.0.17 - November 12, 2020

Fixes

  • Update Session id after reconnection

Version 6.0.16 - October 12, 2020

Fixes

  • Dispatch system:error:expired-token event in application level
  application.on('system:error:expired-token', 'NXM-errors', () => {
    console.log('Token Expired');
  });
  • Handle uncaught exception of conversation:error:invalid-member-state event

Version 6.0.15 - September 21, 2020

Fixes

  • Set SDK default to single preset ICE candidate
  • Remove iceGatherOnlyOneCandidate flag from client configuration

Version 6.0.13 - September 14, 2020

Fixes

  • Update websocket reconnection logic for token expiry

Version 6.0.12 - September 01, 2020

Fixes

  • Update resolved stream on media enabling

Version 6.0.11 - September 01, 2020

Changes

  • Update websocket reconnection logic
  • Fix delay in establishing media connection with one ice candidate sent

Version 6.0.10 - May 04, 2020

Fixes

  • Filter IPv6 Candidates from iceGatherOnlyOneCandidate SDP offer

Version 6.0.9 - March 24, 2020

Fixes

  • Fix handler of rtc:transfer event to refresh conversation

Version 6.0.8 - February 28, 2020

Fixes

  • Fix duplicated webrtc offer sent during IP calling
  • Fix Safari WebRTC dependency

Version 6.0.7 - January 16, 2020

Fixes

  • Fix ANSWERED call status in IP - IP calling
  • Fix docs issues

Changes

  • Improve TypeScript definitions

Version 6.0.6 - November 19, 2019

Fixes

  • Add iceGatherOnlyOneCandidate configuration option and use to define path in ICE gathering process

Version 6.0.5 - November 19, 2019

Fixes

  • Update styling of JSDocs to Docstrap template
  • Change RTCPeerConnection ICE candidates gathering process to send the Session Description Protocol (SDP) offer on first ICE candidate gathered

Version 6.0.4 - November 14, 2019

Fixes

  • Remove remaining audio elements after transferring a call to a new conversation
  • Update conversation.invite() to not include empty user_id or user_name fields within the requests

Version 6.0.3 - October 22, 2019

New

  • Added TypeScript definition files

Changes

  • Added options for customized logging levels in the console of debug, info, warn, or error.
new NexmoClient({
  debug: 'info'
})
  • Moved storage of JWT token from localStorage to NexmoClient configuration object
  • Removed unnecessary files from the NPM release package

Fixes

  • Fixed call statuses order in case of a transfer

Version 6.0.1 - September 27, 2019

Changes

  • Removed media.record() function
  • Removed cache option from SDK, used for storing conversations and events
  • Removed automatic syncing of all individual conversations in login, when sync is lite or full

Version 6.0.0 - September 13, 2019

Breaking Changes

  • Change return value of application.getConversations() to new ConversationsPage object
// iterate through conversations
application
  .getConversations({ page_size: 20 })
  .then((conversations_page) => {
    conversations_page.items.forEach(conversation => {
      render(conversation);
    })
  });
  • Change return value of conversation.getEvents() to new EventsPage object
// iterate through events
conversation
  .getEvents({ event_type: `member:*` })
  .then((events_page) => {
    events_page.items.forEach(event => {
      render(event);
    })
  });
  • Rename method application.callPhone to application.callServer
  • Rename method application.call to application.inAppCall
  • Rename method call.createPhoneCall to call.createServerCall
  • Rename class Call to NXMCall
  • Rename class ConversationClient to NexmoClient
  • Rename class ConversationClientError to NexmoClientError
  • Rename files conversationClient.js and conversationClient.min.js to nexmoClient.js and nexmoClient.min.js
  • Deprecate member:call:state event (use instead member:call:status)
  • Remove automatic login in case of a websocket reconnection and emit the event

New

  • Send and listen for custom event types in a conversation.
//sending a custom event type to a conversation
conversation
  .sendCustomEvent({type: `my_custom_event`, body: { enabled: true }})
  .then((custom_event) => {
    console.log(event.body);
  });
//listening for a custom event type
conversation.on(`my_custom_event`, (from, event) => {
  console.log(event.body);
});
  • Add new PageConfig class for configuring settings for paginated requests
  • Add new Page class to wrap results of paginated requests
  • Add setup of default pagination configuration for conversations and events in ConversationClient initialization
  • Add wild card supported for filtering by event types using :* (for example event_type: member:*)
new NexmoClient({
  conversations_page_config: {
    page_size: 25,
    order: 'asc'
    cursor: 'abc'
  },
  events_page_config: {
    page_size: 50,
    event_type: `member:*`
  }
})
  • Add new ConversationsPage and EventsPage which extend Page class to wrap results of paginated requests for conversations and events
  • Add getNext() and getPrev() methods to ConversationsPage and EventsPage objects to fetch previous and next pages of conversations and events
  • Add conversations_page_last parameter to application object and events_page_last parameter to conversation object for reference to last page retrieved
application.conversations_page_last
  .getNext((conversations_page) => {
    conversations_page.items.forEach(conversation => {
      render(conversation)
    })
  })
conversation.events_page_last
  .getPrev((events_page) => {
    events_page.items.forEach(event => {
      render(event)
    })
  })
  • Add the ability to make an IP-IP call through callServer function
// IP-IP call scenario
application
  .callServer('username', 'app')
  .then((nxmCall) => {
    // console.log(nxmCall);
  });

// IP-PSTN call scenario
application
  .callServer('07400000000')
  .then((nxmCall) => {
    // console.log(nxmCall);
  });

Changes

  • Update reason object to receive reason.reason_text and reason.reason_code fields

Internal changes

  • Rename Event class to NXMEvent
  • Update CAPI requests to REST calls for these events
    • event:delivered
    • text:delivered
    • image:delivered
    • event:seen
    • text:seen
    • image:seen
    • conversation:events
    • audio:play
    • conversation:delete
    • conversation:invite
    • text
    • text:typing:on
    • text:typing:off
    • new:conversation
    • conversation:get
    • user:conversations
    • user:get
    • conversation:join
    • audio:say
    • audio:earmuff:on
    • audio:earmuff:off
    • audio:dtmf
    • audio:record
    • audio:play
    • conversation:member:delete
    • event:delete
    • audio:ringing:start
    • audio:ringing:stop
    • audio:mute:on
    • audio:mute:off
    • image
    • rtc:new
    • rtc:answer
    • rtc:terminate
    • knocking:new
    • knocking:delete

Version 5.3.4 - July 18, 2019

Fixes

  • Custom SDK config object does a deep merge with default config object

Version 5.3.3 - June 29, 2019

Fixes

  • Change digits to digit in the sendDTMF() request method payload.
  • Stream is not being terminated on a call transfer.
  • member:call is not being emitted if media.audio_settings.enabled is false or doesn't exist.

New

  • Set member.callStatus to started when initializing an IP - IP call.
  • Set member.callStatus to ringing when enabling the ringing with media.startRinging().

Internal changes

  • Move stream cleanup from member:left to rtc:hangup in Media module.

Version 5.2.1 - June 12, 2019

New

  • Add the new nexmoGetRequest utility method to make a GET network request directly to CS:
/**
 * Perform a GET network request directly to CS
 *
 * @param {string} url the request url to CS
 * @param {string} data_type the type of data expected back from the request (events, conversations, users)
 * @param {object} [params] network request params
 * @param {string} [params.cursor] cursor parameter to access the next or previous page of a data set
 * @param {number} [params.page_size] the number of resources returned in a single request list
 * @param {string} [params.order] 'asc' or 'desc' ordering of resources (usually based on creation time)
 * @param {string} [params.event_type] the type of event used to filter event requests ('member:joined', 'audio:dtmf', etc)
 *
 * @returns {Promise<XMLHttpRequest.response>} the XMLHttpRequest.response
 * @static
 * @example <caption>Sending a nexmo GET request</caption>
 */
  nexmoGetRequest(url, data_type, params).then((response) => {
    response.body: {},
    response.cursor: {
        prev: '',
        next: '',
        self: ''
    },
    response.page_size: 10
 });
  • Support reason for member:delete, conversation.leave, member.kick, call.hangup and call.reject.
  • Listen for the member:left event with reason:
//listening for member:left with reason
conversation.on('member:left', (member, event) => {
  console.log(event.body.reason);
});

/**
* Reason object format
*
* @param {object} [reason] the reason for kicking out a member
* @param {string} [reason.code] the code of the reason
* @param {string} [reason.text] the description of the reason
*/
  • Add callStatus field in the Member object, defining the status of a call.
  • Emit member:call:status event each time the member.callStatus changes:
conversation.on("member:call:status", (member) => {
   console.log(member.callStatus);
});

Version 5.2.0 - May 30, 2019

New

  • Add the call instance in application.calls map in createCall() function (IP -IP call)

  • Update caller parameter in call object in a PSTN - IP call from unknown to channel.from.number or channel.from.uri if exists

  • Emit the new leg:status:update event each time a member leg status change

/**
  * Conversation listening for leg:status:update events.
  *
  * @event Conversation#leg:status:update
  *
  * @property {Member} member - the member whose leg status changed
  * @property {Event} event - leg:status:update event
  * @param {string} event.cid - the conversation id
  * @param {string} event.body.leg_id - the conversation leg id
  * @param {string} event.body.type - the conversation leg type (phone or app)
  * @param {string} event.body.status - the conversation member leg status
  * @param {Array} event.body.statusHistory - array of previous leg statuses
*/
conversation.on("leg:status:update", (member, event) {
  console.log(member, event);
});
  • Add the the channel.legs field in member events offered by CS
conversation.on(<member_event>, (member, event) {
  console.log(event);
  // member_id: <member_id>,
  // conversation_id: <conversation_id>,
  // ...
  // channel: {
  //  to: {
  //    type: app
  //  },
  //  type: app,
  //  leg_ids: [<leg_id>]
  //  legs : [{ leg_id: <leg_id>, status: <leg_status>}],
  //  leg_settings: {},
  // },
  // state: <state>,
  // leg_ids: []
});

Version 5.1.0 - May 29, 2019

New

  • Send DTMF event to a conversation
  * Send DTMF in a conversation
  *
  * @param {string} digits - the DTMF digit(s) to send
  * @returns {Promise<Event>}
 conversation.media.sendDTMF('digits')
  • Emit new event audio:dtmf 
conversation.on("audio:dtmf",(from, event)=>{
  event.digit // the dtmf digit(s) received
  event.from //id of the user who sent the dtmf
  event.timestamp //timestamp of the event
  event.cid // conversation id the event was sent to
  event.body // additional context about the dtmf
});
  • Set customized audio constraints for IP calls when enabling audio
 conversation.media.enable({
    'audioConstraints': audioConstraints
 })
  * Replaces the stream's audio tracks currently being used as the sender's sources with a new one with new audio constraints
  * @param {object} constraints - audio constraints
  * @returns {Promise<MediaStream>} - Returns the new stream with the updated audio constraints.
  * @example
  * conversation.media.updateAudioConstraints({'autoGainControl': true})
  **/
  • Update audio constraints for existing audio tracks
  conversation.media.updateAudioConstraints(audioConstraints)
 })

Fixes

  • Remove 'this' passed to cache worker event handler

Internal breaking changes

  • Change the media audio parameter from media.audio to media.audio_settings in inviteWithAudio function

Version 5.0.3 - May 23, 2019

Changes

  • Change default behavior of autoPlayAudio in media.enable() from false to true
  • Pass an autoPlayAudio parameter to call.createCall() and call.answer() functions (default is true)

Version 5.0.2 - May 30, 2019

New

  • Delete the image files before sending the image:delete request
  • Attach of audio stream can now be chosen if it will be automatically on or off through media.enable() 
media.enable({
  autoPlayAudio: true | false
})

Changes (internally)

  • Combine the network GET, POST and DELETE requests in one generic function

Version 5.0.1 - April 30, 2019

Fixes

  • Clean up user's media before leaving from an ongoing conversation

Breaking changes

  • Change application.conversations type from Object to Map

Version 4.1.0 - April 26, 2019

Fixes

  • Fixed the bug where the audio stream resolved in media.enable() is causing echo and was not the remote stream
  • Resolve the remote stream pc.ontrack() and not the localStream from getUserMedia

Changes

  • Rename localStream to stream in media.rtcObjects object.

Version 4.0.2 - April 17, 2019

Changes

  • Removed media.rtcNewPromises

New

  • Internal lib dependencies update
  • Added support for Bugsnag error monitoring and reporting tool
 * @class ConversationClient
 *
 * @param {object} param.log_reporter configure log reports for bugsnag tool
 * @param {Boolean} param.log_reporter.enabled=false
 * @param {string} param.log_reporter.bugsnag_key your bugsnag api key / defaults to Vonage api key
  • Updated vscode settings to add empty line (if none) at end of every file upon save
  • Disable the ice candidates trickling in ice connection
  • Wait until most of the candidates to be gathered both for the local and remote side
  • Added new private function editSDPOrder(offer, answer) in rtc_helper.js to reorder the answer SDP when it's needed
  • For rtc connection fail state
    • Disable leg
    • emit new event media:connection:fail 
member.on("media:connection:fail",(connection_details)=>{
  connection_details.rtc_id // my member's call id / leg id
  connection_details.remote_member_id // the id of the Member the stream belongs to
  connection_details.connection_event: // the connection fail event
  connection_details.type // the type of the connection (video or screenshare)
  connection_details.streamIndex // the streamIndex of the specific stream
});
* @event Member#media:connection:fail
*
* @property {number} payload.rtc_id the rtc_id / leg_id
* @property {string} payload.remote_member_id the id of the Member the stream belongs to
* @property {event} payload.connection_event the connection fail event
  • Add new LICENCE file

Breaking changes (internally)

  • Deprecating ice trickling logic with onicecandidate event handler
  • Change the format of member:media event to the new one offered by CS
type: 'member:media',
  from: member.member_id,
  conversation_id: member.conversation_id,
  body: {
    media: member.media,
    channel: member.channel
  }
  • Change the format of member:invited event to the new offered by CS
type: 'member:invited',
  body: {
    media: {
      audio_settings: {
        enabled: false,
        earmuffed: false,
        muted: false
      }
    }
  }

Version 4.0.1 - March 4, 2019

New

  • Select the sync level for the login process
    • full: trigger full sync to include conversations and events
    • lite: trigger partial sync, only conversation objects (empty of events)
    • none: don't sync anything

if the Cache module is enabled the manual fetch of a conversation will store them in internal storage

usage:

  new ConverationClient({'sync':'full'});

Fixes

  • rtcstats:report was duplicating instances in each call
  • remove screenshare https restriction

Breaking changes (internally)

  • Deprecating application.activeStream, now it's part of application.activeStreams
  • Removed the restriction to allow calling media.enable() while a stream is active

Version 4.0.0 - February 1, 2019

Breaking Changes

  • rename SDK stitch to client
  • listening for media:stream:* now gives streamIndex instead of index for consistency with the internal rtcObjects
 * @event Member#media:stream:on
 *
 * @property {number} payload.streamIndex the index number of this stream
 * @property {number} [payload.rtc_id] the rtc_id / leg_id
 * @property {string} [payload.remote_member_id] the id of the Member the stream belongs to
 * @property {string} [payload.name] the stream's display name
 * @property {MediaStream} payload.stream the stream that is activated
 * @property {boolean} [payload.video_mute] if the video is hidden
 * @property {boolean} [payload.audio_mute] if the audio is muted

New

  • Screen Share Source ID can now be specified when invoking media.enable()
Improve this page