- AppHandler::render now receives the current &Frame; its default
implementation renders the whole scene automatically via
app.render_scene(frame.view()) (Option A). Users can simply not
implement render for full auto-rendering.
- Add Renderer::render_scene: batch-renders every scene entity in a
single render pass. Factored per-mesh draw logic into a private
draw_entity helper shared with Renderer::render.
- Add App::render_scene(view) delegating to the Renderer.
- Fill simple.rs with a real quad (mesh/material/entity) without
importing wgpu; the scene now auto-renders via the trait default.
DRAFT.md now lives in docs/. Confirmed the pending design decisions:
render() auto-renders the scene by default (Option A, alternative B removed),
and simple.rs uses app.renderer.device()/format() via the library API
without importing wgpu.
Codifies the systematic documentation of public items with the API docs
in mind, based on the incidents from the last session: backticks around
every type (Option<Self>, Arc<Mesh>, [f32; 3]) to avoid unresolved-link
and unclosed-HTML-tag warnings, one doc comment per public item, the
#![warn(missing_docs)] policy, verifying '0 warnings' via cargo doc
before committing, and doc-test conventions (rust vs ignore blocks).
Applies to wsg-lib development from now on.
- Wrap in backticks every bare type in rustdoc comments (vertex.rs,
frame.rs, pipeline_cache.rs, material.rs, mesh.rs, scene.rs, error.rs)
so rustdoc no longer misreads them as intra-doc links or HTML tags.
- Add a missing doc comment on the public Frame struct.
- Add #![warn(missing_docs)] to the crate root so unevidenced public
items are surfaced going forward.
cargo doc --no-deps now generates with zero warnings.
Note in the ROADMAP decision table that the swapchain stays on
PresentMode::Fifo (double-buffered vsync, latency 2) for now. Mailbox
(triple buffering) remains a future opt-in and Immediate stays reserved
for offscreen; present mode is revisited only when the GPU-driven
pipeline (Phase 3) lands. Not blocking for Phases 1-2.
Clarify for each docs/tech document whether it describes the current
(implemented) or target (planned) architecture, and reconcile with the
readme/roadmap source of truth:
- ARCHI_APP, ARCHI_CPU_GPU, ARCHI_RENDU: mark status=target and add a
banner stating the GPU-driven two-pass pipeline, persistent VRAM buffers
and scene auto-render are not implemented (they are ROADMAP Phase 3 /
README roadmap items 1-3). render() still cannot draw the scene.
- ARCHI_ARENES: status=target/deferred. String IDs are the current design;
slotmap generational handles are the deferred 'typed handles' step and
the slotmap dependency is currently absent (supersedes the retired dep).
- FRAME_LOOP: stays status=current; correct the described API flow (Frame
based via get_next_frame/Renderer::present, plus the low-level
begin_frame/end_frame variant) and move the single-buffer VRAM note to
target.
- README: label each tech doc as current or target in the Documentation
section.
- Replace the broken App::new() (which does not exist) with AppBuilder.
- Strip it down to the PLAN's ~15-line target: no explicit winit/wgpu,
AppHandler render() left empty (scene rendering still on the roadmap).
- Verified: builds, opens a black window and runs the update->render->present
loop until closed.
- Update README/PLAN wording now that the simple example compiles (it still
does not draw a scene).
- README stays the source of truth for current state; drive/item-5 now
reflects the settled decision: String IDs for the MVP, slotmap deferred.
- ROADMAP gains a 'point de depart' (present-state) block, marks glam done,
removes the slotmap Phase-1 mandates, and documents the String-ID decision
plus the deferred 'typed handles' step.
- PLAN corrects stale [X] checkmarks (App render() cannot draw, Scene
rendering not automated, simple example does not compile) and notes its
verification checklist against current reality.
- Remove the unused slotmap direct dependency from wsg-lib (package remains
in Cargo.lock only as a transitive dep of glow).