Repository navigation
Desktop UI Architecture
Section: Desktop App · Mixing-and-Export · Home · Timeline-and-Waveform-Editor
The desktop UI is a cross-platform Avalonia / .NET 10 shell for the Trackdub workstation: media ingest, transcript editing, translation review, voice synthesis, and export. It depends on Application, Composition, Domain, Licensing, Media.Playback, and Sdk, and stays isolated from direct inference and persistence implementation.
The UI uses ReactiveUI for live async flows (pipeline progress, playback) and CommunityToolkit.Mvvm for ordinary UI state and commands.
MVVM with strict state ownership rules. Trackdub.App.Avalonia multi-targets net10.0 (Linux/macOS) and net10.0-windows10.0.19041.0 (Windows). It must not contain pipeline orchestration, inference code, or SQL.
graph TD
App[Trackdub.App.Avalonia] --> AppLayer[Trackdub.Application]
App --> Comp[Trackdub.Composition]
App --> Domain[Trackdub.Domain]
App --> Playback[Trackdub.Media.Playback]
App --> Sdk[Trackdub.Sdk]
App --> Licensing[Trackdub.Licensing]
subgraph UI_Layer_Responsibilities
Views[Views / .axaml]
VMs[ViewModels]
Conv[Converters]
end
| Toolkit | Usage | Examples |
|---|---|---|
| ReactiveUI | Live/reactive data flows and async events. | Pipeline stage status, operation progress, playback state. |
| CommunityToolkit.Mvvm | Standard UI state, toggles, and simple commands. | Shell toggles, settings fields, form inputs. |
Do not mix patterns on the same property.
Shell chrome hosts transport, timeline, side panels, and stage controls. Pipeline stages are user-triggered (per-stage Run and global run). Media load alone does not start the speech pipeline.
Waveform right-click is for diarization edit (split/merge/delete). Segment retranscribe/translate/redub lives on the segment context menu or expanded row.
Trackdub Engineering Wiki · Trackdub-gated · Trackdub · Contributing
Overview
Architecture
Pipeline
- Pipeline-Orchestration
- Media-Ingest-and-Preparation
- Transcription-and-VAD
- Diarization-and-Speaker-Management
- Translation-Engine
- Text-to-Speech
- Lip-Sync-and-Synthesis
- Mixing-and-Export
Desktop App
AI & Inference
- Inference-Routing-and-Planning
- ONNX-Execution-Providers
- Model-Manifests-and-Governance
- Hardware-Starter-Packs
Media & Audio
Data
Cloud & SaaS
- Cloud-API-and-Job-Management
- Web-Dashboard
- Billing-and-Quota-Management
- Webhook-Delivery-System
- Cloud-Infrastructure-Deployment
Development
Links