<!-- This is an auto-generated comment: release notes by coderabbit.ai -->
## Summary by CodeRabbit
- **New Features**
- Documents now automatically track and display "created by" and "updated by" user information.
- Document creation and update timestamps are now managed and shown more accurately.
- Workspace and document metadata (name, avatar) updates are more responsive and reliable.
- Document creation supports middleware for customizing properties and behavior.
- **Improvements**
- Simplified and unified event handling for document list updates, reducing redundant event subscriptions.
- Enhanced integration of editor and theme settings into the document creation process.
- Explicit Yjs document initialization for improved workspace stability and reliability.
- Consolidated journal-related metadata display in document icons and titles for clarity.
- **Bug Fixes**
- Fixed inconsistencies in how workspace and document names are set and updated.
- Improved accuracy of "last updated" indicators by handling timestamps automatically.
- **Refactor**
- Removed deprecated event subjects and direct metadata manipulation in favor of more robust, reactive patterns.
- Streamlined document creation logic across various features (quick search, journal, recording, etc.).
- Simplified user avatar display components and removed cloud metadata dependencies.
- Removed legacy editor setting and theme service dependencies from multiple modules.
- **Chores**
- Updated internal APIs and interfaces to support new metadata and event handling mechanisms.
- Cleaned up unused code and dependencies related to editor settings and theme services.
- Skipped flaky end-to-end test to improve test suite stability.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AF-2572
<!-- This is an auto-generated comment: release notes by coderabbit.ai -->
## Summary by CodeRabbit
- **New Features**
- Introduced a Meetings welcome page with a beta disclaimer and "Get Started" flow.
- Added separate toggles for AI auto summary and AI auto todo list in meeting settings.
- Added "Beta" labels to relevant settings and sidebar items for clearer feature status.
- Enhanced settings UI with improved headers, subtitles, and new styling.
- **Improvements**
- Meeting settings now allow independent control over AI-generated summaries and todo lists.
- Updated internationalization to support new meeting and AI transcription features, including richer prompts and hints.
- Refined logic for enabling meeting recording, including improved permission handling.
- Simplified transcription logic to rely solely on AI enablement flag.
- **Bug Fixes**
- Fixed display and control of meeting settings based on beta disclaimer acceptance.
- **Chores**
- Updated localization files and completeness percentages for several languages.
- Removed deprecated feature flag for enabling meetings.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
fix AF-2460, AF-2463
When recording is started, we start polling the pending raw buffers that are waiting for encoding. The buffers are determined by the cursor of the original raw buffer file. When recording is stopped, we will flush the pending buffers and wrap the encoded chunks into WebM.
```mermaid
sequenceDiagram
participant App as App/UI
participant RecordingFeature as Recording Feature
participant StateMachine as State Machine
participant FileSystem as File System
participant StreamEncoder as Stream Encoder
participant OpusEncoder as Opus Encoder
participant WebM as WebM Muxer
Note over App,WebM: Recording Start Flow
App->>RecordingFeature: startRecording()
RecordingFeature->>StateMachine: dispatch(START_RECORDING)
StateMachine-->>RecordingFeature: status: 'recording'
RecordingFeature->>StreamEncoder: createStreamEncoder(id, {sampleRate, channels})
Note over App,WebM: Streaming Flow
loop Audio Data Streaming
RecordingFeature->>FileSystem: Write raw audio chunks to .raw file
StreamEncoder->>FileSystem: Poll raw audio data
FileSystem-->>StreamEncoder: Raw audio chunks
StreamEncoder->>OpusEncoder: Encode chunks
OpusEncoder-->>StreamEncoder: Encoded Opus frames
end
Note over App,WebM: Recording Stop Flow
App->>RecordingFeature: stopRecording()
RecordingFeature->>StateMachine: dispatch(STOP_RECORDING)
StateMachine-->>RecordingFeature: status: 'stopped'
StreamEncoder->>OpusEncoder: flush()
StreamEncoder->>WebM: muxToWebM(encodedChunks)
WebM-->>RecordingFeature: WebM buffer
RecordingFeature->>FileSystem: Save as .opus file
RecordingFeature->>StateMachine: dispatch(SAVE_RECORDING)
```
- Added a check to verify if AI is enabled before attempting to transcribe meeting recordings
- Improved error handling for empty recordings
- Fixed the recording timeout logic to ensure it only stops the correct recording session
Added a recording popup UI for the audio recording feature in the desktop app, improving the user experience when capturing audio from applications.
### What changed?
- Created a new popup window system for displaying recording controls
- Added a dedicated recording UI with start/stop controls and status indicators
- Moved audio encoding logic from the main app to a dedicated module
- Implemented smooth animations for popup appearance/disappearance
- Updated the recording workflow to show visual feedback during recording process
- Added internationalization support for recording-related text
- Modified the recording status flow to include new states: new, recording, stopped, ready
fix AF-2340
AudioMedia entity for loading & controlling a single audio media
AudioMediaManagerService: Global audio state synchronization across tabs
AudioAttachmentService + AudioAttachmentBlock for manipulating AttachmentBlock in affine - e.g., filling transcription (using mock endpoint for now)
Added AudioBlock + AudioPlayer for rendering audio block in affine (new transcription block whose renderer is provided in affine)
fix AF-2292
fix AF-2337
A listening tappable app's info should inherit from its group process's name/icon. However the group process may not be listed as a tappable application.
