How we prioritise our objectives
We use a Mural board with a priority radar to discuss our shared priorities across different topics. Cards that are of primary importance are placed in the radar's innermost circle.
Summary
Clarify and Define Separation of Concerns
Features
Ownership
- Who owns data value resolving? (USD?)
- Who owns usd<->mtlx synchronization? (USD?)
- Who owns the mtlx->usd conversion?
- Who owns the usd->mtlx conversion?
- Who owns validation (USD?)
- Who owns reference definitions, who owns implementations (MaterialX?), including USDPreviewSurface?
- Who owns code generation access.
- Who decides what is common material metadata? There's not a 1:1 mapping between USDShade and MaterialX so how can we drive standardisation?
Alignment and Feature Parity
Lossless interop
- Applicable USD limitations
- Assignment expressions
- Round trip:
- Mtlx -> USDShade -> Mtlx
- Mtlx -> USDShade -> *USDShade (e.g. shot override) -> *Mtlx (with the changes)
Other
- Dependency tracking, compatibility, and versioning
- Cross referenced documentation
- End-to-end colour management
Fast and Robust Hydra Rendering
- Improve MaterialX - UsdShade updates (both topology and inputs)
- USDShade with MaterialX source code
- Translation / ShaderGen independent from Hydra
Fit for Purpose - Extensibility
- Strive for standardisation but empower diverse workflows
- Standardise new shading models
- Support workflows relying on custom USD schemas and MaterialX node definitions
Guide Community
- Promote MaterialX as a material description in USD
- Curate test and validation assets
- Promote and document best practices
- MaterialX blackbox references, with MaterialX overrides expressed in USD?
- General pipeline considerations