amazon-chime-sdk-js - npm Package Versions

[2.18.0] - 2021-09-22

Added

• Add events meetingReconnected, signalingDropped and receivingAudioDropped to eventDidReceive by publishing them as stand alone events. Currently, these events were only included in the meeting history attribute when a meeting event is published.
• Added support for skipping full SDP renegotiations when switching simulcast streams. This will result in less freezing when switching between layers in response to a network event as done in VideoPriorityBasedPolicy. This will have no impact if not using simulcast.
• Add link to SIP Media Application examples in README.

Removed

Fixed

• Add safeguard for Nscale policy in case we increase to more than 25 videos.

Changed

• Move toLowerCasePropertyNames inside Utils.ts and add test coverage.
• Reduced uplink resubscription when only stream encoding is changed by adding bypassing path.
• The browser demo now offers a configuration menu on each video tile. This menu replaces the 'Pin' button, which previously set the priority of the corresponding remote video to 1, and then rest to 2. The new configuration menu allows the user to specify the desired video quality and priority, which will be respected by simulcast and priority downlink policies. This is useful for testing or to demonstrate the behavior of those policies.
• Switched to using Web Components for video tiles in the browser demo.
• Migrate SauceLabs mobile tests to new api.

[2.17.0] - 2021-09-08

Added

• Add audioInputMuteStateChanged to the DeviceChangeObserver interface. This is called whenever the device is changed or is muted or unmuted, allowing applications to adapt to OS-level mute state for input devices.
• Added Android WebView Sample UI test to workflow.
• Add a new optional API getVideoTileForAttendeeId in VideoTileController and raise the tileWillBePausedByDownlinkPolicy event for empty video tiles.
• Amazon Voice Focus will now trigger the onCPUWarning method on the VoiceFocusTransformDeviceDelegate when using the default inline execution mode. This will be called when the browser fails to schedule the worklet in a timely fashion (e.g., when the meeting code is running in an iframe /subframe) or when changes in the CPU environment (e.g., thermal throttling) cause the worklet to take too long for each audio render quantum.
• Amazon Voice Focus will now trigger the voiceFocusInsufficientResources method on the VoiceFocusTransformDeviceDelegate when using the default inline execution mode. This will be called when the browser fails to schedule the worklet in a timely fashion (e.g., when the meeting code is running in an iframe /subframe) or when changes in the CPU environment (e.g., thermal throttling) cause the worklet to take too long for each audio render quantum.
• Add retry logic in FAQs.
• The Amazon Voice Focus support check, VoiceFocusDeviceTransformer.isSupported, now warns to the logger when run in an iframe, and can be configured to fail in that case.
• Add documentation in Video Processing APIs on how to add filters in the preview window.

Changed

• Clarify why not use default downlink policy with simulcast.
• Corrected argument isUnifiedPlan in withBandwidthRestriction to isFirefox. Also marked as deprecated since we no longer use it.
• Update data message limit in the API Overview guide.
• Do not trigger real time attendee presence event for local attendee if they are not appear in audio info frame during reconnection.
• Update startVideoPreviewForVideoInput to support filters in the preview window.
• Update browser demo to showcase preview filter capability.

Removed

Fixed

• Fix priority-based downlink policy to not unpaused tiles that are not paused by the policy.
• Fix empty video tiles when using priority-based downlink policy.
• Fix simulcast guide that adaptive probe downlink policy is not enabled by default.
• Fix a link format in simulcast guide.
• No longer put useless 'pin' and 'pause' buttons on local tile in demo.
• Choose a null device or media stream without a deviceId without first listing devices no longer logs Device cache is not populated.

[2.16.1] - 2021-08-23

Fixed

• Fix default priority downlink policy to update default preference correctly.

[2.16.0] - 2021-08-17

Added

• Add RealtimeSubscribeToAttendeeIdPresenceCallback type for realtimeSubscribeToAttendeeIdPresence callback to document the callback parameters.
• Added support for Android WebView
• Add a SignalClientEvent check in SubscribeAndReceiveSubscribeAckTask to immediately cancel the task when websocket connection is terminated.

Changed

• Update the default behavior of NScale video uplink bandwidth policy to scale down resolution based on the number of videos.

Removed

Fixed

• Fix race condition in Safari when disconnect and connect stream from video element.

[2.15.0] - 2021-08-04

Added

• Supports integration with Amazon Transcribe and Amazon Transcribe Medical for live transcription. The Amazon Chime Service uses its active talker algorithm to select the top two active talkers, and sends their audio to Amazon Transcribe (or Amazon Transcribe Medical) in your AWS account. User-attributed transcriptions are then sent directly to every meeting attendee via data messages. Use transcriptions to overlay subtitles, build a transcript, or perform real-time content analysis. For more information, visit the live transcription guide.
• [Demo] Add live transcription functionality. You will need to have a serverless deployment to create new AWS Lambda endpoints for live transcription. Follow the live transcription guide to create necessary service-linked role so that the demo app can call Amazon Transcribe and Amazon Transcribe Medical on your behalf.
• Exposed Amazon Voice Focus model complexity as a type in order to support showcasing complexity limitation in the meeting demo.
• Packet-per-second (PPS) logging is now enabled in the meeting demo by default. If the browser sends an incorrect packet rate, this will be logged as an error in the console.
• Add a warning log in InMemoryJSONEventBuffer's send function when retrying starts.

Changed

• Update InMemoryJSONEventBuffer to retry with backoff.

Removed

Fixed

• Stop activeDevice video track before selecting a new device to prevent NotReadableError when calling getUserMedia for a new video input device.
• Fix priority-based downlink policy default behavior.
• Fix client event ingestion guide rendering in typedoc.

[2.14.0] - 2021-07-23

Added

• Added VideoPriorityBasedPolicyConfig to control video downlink policy with network event response and recovery delays. Check User Guide for Priority-based Downlink Policy for more information.
• Amazon Chime SDK Project Board Overview and Guide.
• Added 25 video tile support for demo app.

Changed

• [Documentation] Update priority based downlink policy guide.
• Lookup groupId from device cache instead of directly from media stream when selecting input device.
• [Documentation] Update documentation for 25 video tiles.

Removed

Fixed

• Improve the meeting event guide.
• Fixed Project Board guide with correct community template link.
• Updated Amazon Voice Focus integration guide to reflect recent Safari versions.
• Import current Amazon Voice Focus code, which ensures that stereo inputs are downmixed to mono.

[2.13.0] - 2021-06-29

Added

Changed

Removed

Fixed

• Improve the meeting event guide

[2.12.0] - 2021-06-23

Added

• [Documentation] Add documentation for getObservableVideoMetrics.
• [Documentation] Update FAQ and public documentation to add more information on SignalingBadRequest related error codes.
• [Documentation] Rephrase the terms in the status code documentations.

Changed

• Bump maxVideos limit to 25

Removed

Fixed

• Pre-started signaling connections no longer cause a delay in joining if the user takes more than a minute to join the meeting.
• Fix choosing input device API when passing in a media stream.

[2.11.0] - 2021-06-04

Added

• Bind tileController during the initialization of DefaultAudioVideoController for VideoPriorityBasedPolicy.
• Add more debug logging for choose input device.
• Add the meeting and device error sections in the meeting-event guide.
• Add a forceUpdate parameter to use when listing devices. In some cases, builders need to delay the triggering of permission dialogs, e.g., when joining a meeting in view-only mode, and then later be able to trigger a permission prompt in order to show device labels. This parameter allows cached device labels to be forcibly discarded and recomputed after the device label trigger is run.

Changed

• Log error instead of throwing error if the signaling client is not ready to send data message.
• Now when setDeviceLabelTrigger is called, if the deviceInfoCache contains a device with no label, deviceInfoCache will be cleared.

Removed

• Remove deprecated unwired webrtc constraints from device controller and peer connection construction.
• Removed unnecessary restriction on VideoPriorityBasedPolicy to always subscribe to at least one stream.

Fixed

• Fixed missing upstream video metrics for Firefox browsers.
• Fix build script to run on Windows by specifying ruby when running ruby scripts and rimraf to remove folder.

[2.10.0] - 2021-05-19

Added

• Add new message MeetingSessionStatusCode AudioAttendeeRemoved to handle the new audio server status code 411.
• Add support for WKWebView on iOS.
• Output a warning message when the volume adapter cleans up the self-attendee after reconnection.
• Add FAQ for more information on AudioJoinFromAnotherDevice meeting session status code.
• Add downstream audio webrtc metrics in observableMetricSpec.
• Add getObservableVideoMetrics and in ClientMetricReport to expose video stream metrics in webrtc.
• Update SignalingProtocol with optional video metric fields.

Changed

• Update guide for priority based downlink policy.
• Bump version for lodash, y18n, and ssri dependencies.
• Mark getObservableVideoMetrics optional in ClientMetricReprt and videoStreamIndex and selfAttendeeId optional in DefaultClientMetricReport.

Removed

Fixed

• Do not start local video tile if there is no stream for content share.

• Media streams are no longer discarded during reconnects. This fixes an issue where initial signaling connection failures could cause a client to be unable to join a meeting with audio if Web Audio were enabled.