Resolve curve intersections through cadkernel, retain empty exact results, gather analytic candidates independently of tessellation, and bound material uploads. Preserve large MINSERT data while limiting rendering work; remove obsolete review narratives.
Highlighting a whole-drawing selection took 299 ms, and it runs when the
selection changes — directly between the click that completes a selection and
seeing it.
Most of that was asking the same question twice. Each wire went through the
circle and ellipse extractors at its draw depth to build the analytic
instances, then through both again at depth 0.0 purely to ask *whether* it is
a circle or an ellipse, allocating a vector each time and dropping it.
Neither extractor's `Some`/`None` answer depends on the depth: every
`return None` in them turns on the geometry — which `TangentGeom` variant,
the radius, the axis lengths, the parameters — and the depth is only written
into the instance, never read to decide. So the extraction that builds the
instances already answers the classification. One pass now does both, and
fans out across cores, since it reads a wire and writes nothing shared. That
took 173 ms of shape extraction and splitting down to 19 ms, and the whole
highlight from 299 ms to 150 ms.
Gathering the highlighted wires also cloned each handle's slot vector and
sorted the copy — one allocation and one sort per selected entity. The slots
are already ascending: both builders walk the wires in order and push as they
go, and `patch_handle_index` shifts a handle's slots by a constant, which
preserves it. A `debug_assert` on the slots actually read holds that in
place. The compat path of `from_highlight_refs` emitted its instances
sequentially while the native path beside it fans out; it now matches.
Nothing renders in a test, so a wrong bucket in the classification would
surface only as circles drawn wrong when selected. Two tests cover it: each
wire lands where the old predicate put it, and the colour override — always
for hover, and the configured tint or `SELECTED` for a selection — reaches
the instances.
`wire-highlight-detail` reports the phases: gathering, classifying,
uploading the analytic instances, the regular wires and the block ones.
Stage 0 of the roadmap asks for a memory baseline before any more
optimisation, and explicitly not from RSS: the allocator and the driver
both retain high-water memory that RSS cannot tell apart from live
resources.
`gpu-live` counts from the sizes this renderer allocated, in the
categories large enough to decide where the next fix goes:
slots total shadow targets text_atlas wire_arena block_geometry
It is partial by construction and says so, rather than presenting itself
as a total. Hatches, images, meshes and Face3D are not counted yet.
Three of the numbers needed a size that was not reachable before: the
block-geometry chunk records its bytes at build because in storage mode
the buffer belongs to the bind group, the arena exposes its allocation
rather than its occupied prefix — headroom is what a budget has to answer
for — and the text atlas reports its `R8` texel count.
Emitted only when a number moves. `prepare` runs thousands of times in a
session and these change a handful of times, so a per-frame line would
bury the log it is meant to explain.
This answers the question the previous fix left open: whether the arena's
52 MiB per slot, the 16 MiB shadow map every slot allocates whether or not
shadows are on, or the render targets dominate once a paper layout has
four or five viewports on screen.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
After an out-of-memory the viewport stopped updating for the rest of the
session — both Model and Layout frozen, the UI still responsive. The log
shows why: `block-wire-geometry` prints *after* the errors it provoked.
The failed uploads were cached.
A rejected allocation still yields a `Buffer`. It is invalid, every later
use reports again, and it is indistinguishable from a good one at the
cache key. So the entry keeps hitting, nothing ever rebuilds, and the
frame stays empty even once memory frees up. Sharing `block_geometry`
across slots in the previous commit made that worse, not better: one
poisoned entry now serves every viewport.
The device's uncaptured-error counter was already there, private to the
handler that throttles the log. It is now readable, and the three caches
that can retain a bad upload — block definition geometry, the shared wire
buffer cache, and the arena's content id — snapshot it across their build
and decline to record what it condemns. The degraded frame still draws;
the next frame tries again.
Best effort by construction: wgpu surfaces some errors asynchronously, so
a moved counter means something failed, while an unmoved one is not proof
that nothing did. That asymmetry is the right way round here — believing
a bad buffer costs a frozen viewport, rebuilding a good one costs a frame.
`block-wire-geometry` gains `poisoned=` so a run says whether this
happened rather than leaving it to be inferred.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The layout-switch OOM names its buffer: `Buffer with 'block_wire.vertices'
label is invalid`, right after `uncaptured wgpu error #1 … Out of Memory`.
That buffer is the largest thing this renderer allocates, and five sixths
of it is duplication.
A segment reaches the shader once per corner of its quad, and the packed
path satisfies that by writing the whole 60-byte record six times —
360 bytes a segment — with `vs_main` recovering the corner from
`vertex_index % 6`. Six identical copies, on a 2 GB card.
`block_wire_storage.wgsl` puts the segments in a read-only storage buffer
indexed by `vertex_index / 6` instead, so each record is stored once. The
geometry vertex buffer disappears; the placements move to slot 0 and the
segments arrive through the definition's bind group, which becomes
per-chunk because each chunk binds its own slice.
This is the pattern `wire_indexed.wgsl` already uses for per-wire
constants, down to the capability gate: `supports_wire_storage` asks for
exactly one vertex-stage storage binding, which is exactly what this
needs. Where it is false — WebGL, the forced compatibility renderer — the
packed shader and its vertex buffer are built as before.
`Segment` is declared with scalar members rather than `vec3<f32>`, which
would align to 16 in storage and reintroduce the padding this removes. A
test validates the shader through naga and asserts the struct's stride
equals `size_of::<BlockWireVertex>()`, because a layout mismatch would
otherwise surface only as garbage geometry on a real device.
`block-wire-geometry` reports the bytes uploaded and what the packed path
would have taken, so the saving is measured rather than assumed.
The cache value becomes `Arc<Vec<BlockGeometryChunk>>` while keeping the
translation-aware key from #1006 — both modes still have to invalidate
when the base instance moves, because the records hold its world
coordinates either way. `const_bgl` and the mode are decided together
when the pipeline is built, so they travel as one `BlockWireTarget`
rather than pushing `from_wires` past the argument limit.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Invalidate cached world-space block vertices when the base translation changes. Share the regular/mesh partition between full uploads and patches so instanced blocks are uploaded only once. Cover both paths with CPU and GPU regression checks.
Adding one entity to a drawing rebuilt every block group's geometry from
scratch. Measured on a 1.17 M-entity drawing with 11 380 block wires, the
per-edit upload split as:
upload-split regular=80.4ms block=230.3ms
The block half is three quarters of it, and every byte of it was
identical to the previous frame: a new line is not a block reference.
`BlockWireGpu::from_wires` re-emitted each group's segments, re-expanded
them six-fold and re-uploaded the vertex buffers anyway.
That work depends only on the block definition being expanded, which
already has a stable identity: `render_instance.source_id`, minted once
when the block cache expands a definition and unchanged while that
expansion stays cached. Cache the vertex chunks and constants under
`(source_id, mesh_edge, colour)` -- colour is in the key because a
highlight tint changes the constants while leaving the vertices alone.
Only the per-instance placements are rebuilt, and they must be: each
carries its own draw depth and the depth map shifts on every add or
remove. They are also small, a few thousand placements against millions
of vertices.
Entries not seen in a build are dropped, so a block-cache rebuild that
mints fresh source ids does not strand old buffers on the GPU.
Measured: the block half falls from 230 ms to 7.4 ms, and the whole
per-edit upload from ~350 ms to ~118 ms.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The previous commit converted the allocations that scale with the
drawing and deliberately left the small fixed-size ones -- uniforms, the
viewcube's static geometry, per-image and per-wipeout quads -- reasoning
that an allocation failure was not a realistic outcome for a few dozen
bytes.
That reasoning was wrong, and the log says so precisely. On the
reproducer the vertex buffers degraded exactly as intended:
[gpu] uncaptured ... Out of Memory
[gpu] uncaptured ... In Queue::write_buffer
Buffer with 'block_wire.vertices' label is invalid
... errors #4#8#16#32#64#128, session kept alive
and then the process died here:
thread 'main' panicked at wgpu_core.rs:2253:
Error in Buffer::get_mapped_range: Validation Error
Caused by: Buffer with 'block_wire.const' label is invalid
'block_wire.const' is a single uniform struct. Once the device is out of
memory every allocation fails, including a 32-byte one, and
create_buffer_init maps it and panics. Size was never the relevant
property -- whether the path maps is.
So every remaining allocation goes through gpu_upload too: block-wire
const, the viewcube's seven buffers, the blit and clip-boundary buffers,
image instances/params/vertices, mesh material params and surface and the
empty index stub, the texture-hatch uniform, and all five wipeout
buffers. `grep -r create_buffer_init src` now finds only a comment, and
no mapped_at_creation:true remains outside gpu_upload's own docs.
`queue` is threaded through the last holdouts: upload_wipeouts,
upload_clip_boundary, MeshBatchStubs::new and
create_material_bind_group_from_resources.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The crash in #358 was not the out-of-memory itself. wgpu reports OOM as
recoverable and the app already installs a handler that logs and keeps
the session alive. What killed the process was what happened next:
create_buffer_init maps the new buffer unconditionally, and mapping the
invalid buffer an OOM device returns panics. That panic unwinds through
iced's main-thread redraw, trips the wgpu-hal swapchain assertion in a
destructor, and becomes a non-unwinding abort -- no chance to autosave.
Every allocation whose size scales with the drawing now goes through
src/scene/pipeline/gpu_upload.rs, which creates the buffer unmapped and
fills it with Queue::write_buffer. A failed allocation then surfaces as a
validation error on the write, which reaches install_gpu_error_handler
and degrades the frame -- exactly what that handler's comment promised:
"A bad frame must degrade, not end the session."
Converted: block-wire vertices and instances, the wire const storage
buffer, all three wire-arena allocators, face3d vertices and instances,
block-text instances and text vertices, both hatch backends, and the
silhouette buffers. mesh_gpu already allocated this way, which is why the
mesh path never produced this panic; the rest now match it.
No mapped_at_creation:true remains in the renderer. The create_buffer_init
calls left are fixed-size or bounded-small -- uniforms, the viewcube's
static geometry, per-image and per-wipeout quads -- where an allocation
failure is not a realistic outcome.
Two trade-offs worth stating:
* write_buffer stages through a copy, which mapping at creation avoided.
Chunking bounds that transient cost, and the measured block-wire
upload got faster with chunking, not slower.
* `queue` had to be threaded through the wire, face3d, text and hatch
builders and their callers. Mechanical, but it widens the diff.
This touches every upload path, so it needs a visual check against a real
drawing before it goes upstream.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
max_buffer_size is an API validation limit, not physical VRAM. NVIDIA
reports several GB of it on a 2 GB card, so sizing a chunk from it
produces buffers larger than the whole device. That is the same bug as
#203 (mesh_gpu), #358 (face3d_gpu) and the block-wire vertex buffer:
rediscovered three times because each upload path spelled the budget out
for itself.
mesh_gpu.rs was the only site that clamped the result. Move that clamp
into src/scene/pipeline/gpu_budget.rs and route every allocating path
through it, so a new upload path cannot reintroduce the mistake by
copying one of the unclamped expressions.
Five sites had no clamp at all: text_gpu (block-text instances),
wire_gpu (block-wire instances), and face3d_gpu (vertices, instances,
and upload_chunks).
Three behaviour changes worth noting:
* Silhouette chunks were capped at a private 64 MiB constant and now
use the shared 32 MiB budget.
* The storage-hatch chunk limit was the minimum of two device limits,
both of them validation figures; it is now also capped by the
budget. The `*_fits` predicates next to it are genuine "would wgpu
reject this" checks and are left alone.
* mesh_gpu's instanced path used the unclamped ceiling to decide
whether a mesh fits in a single buffer, which would still admit a
multi-GB allocation. It now uses the clamped budget, so a few more
meshes take the chunked path.
OCS_GPU_CHUNK_MIB overrides the cap. Raising it above the real VRAM is
the cheapest way to force an allocation failure and check that the
renderer degrades instead of aborting.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
BlockWireGpu::from_wires built one unbounded vertex buffer per
(source_id, mesh_edge) group. On a block-heavy drawing that reached
~505 MiB in a single create_buffer_init, which fails on a 2 GB card;
wgpu then returns an invalid buffer and the unchecked
get_mapped_range_mut() inside create_buffer_init panics.
The instance buffer in the same function was already chunked. Convert
the vertex buffer the same way, matching the solid-mesh chunking in
mesh_gpu.rs (#203) and the fill chunking in face3d_gpu.rs (#358),
including mesh_gpu.rs's clamp: max_buffer_size is an API validation
limit, not physical VRAM, and NVIDIA reports several GB on a 2 GB card.
Chunk sizes are rounded down to a multiple of 6 so a segment's quad
never straddles two buffers -- the shader derives its corner from
vertex_index % 6 and every draw restarts that counter at zero.
Instance buffers are hoisted out so they are built once and reused
across vertex chunks instead of re-uploaded per chunk.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Share source buffers across block placements and render nested content with GPU instance transforms.\n\nChunk silhouettes and compact mesh material data so large drawings stay under device buffer limits and remain responsive.
Pin Iced 0.15 so dialogs can measure content before distributing the resolved frame across child panels. Keep the web plugin notice compact and non-resizable.\n\nCloses #582
Keep storage fast paths on capable adapters and use packed or texture fallbacks only when device limits require them.\n\nCompact mesh-edge varyings to avoid WebGL shader rejection.
Avoid full document, cache, and GPU rebuilds for entity deltas so add, delete, history, selection, and live polyline edits stay responsive on large drawings.
Xref merge now inserts entities in the source block's chain order and
carries SortEntitiesTable objects across (remapped to merged handles).
draw_depth_map values become [depth, half] so exploded block/xref fills
and wide-polyline band wires both compose child depth as
insert_depth + in-block rank x half, recursing into nested inserts with
the same divisor - bands and fills from one block interleave by the
block's internal order instead of stacking at the insert's single depth.
Band batches split per rank via WireModel.depth_override; thin wires
stay merged. The wire arena falls back to a full rebuild when a slab
carries mixed depths so refreshes don't flatten band overrides.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The web wire pipeline (fat-instance wire.wgsl) had grown to 17 vertex
attributes as the width-band and taper fields were added; WebGL2 and
WebGPU cap vertex attributes at 16, so pipeline creation failed and
the web viewport drew no lines at all — UI and SDF text still worked,
which is why it looked like "the app stopped working" rather than a
crash (#414). Scalars now ride in packed vec4/vec2 attributes (10
locations total); the instance struct's memory layout is unchanged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A wide polyline with per-vertex varying widths rendered at its widest edge
(uniform-width ones were already exact). Draw the true taper instead.
`WireModel.taper_widths` carries a per-point full band width. The wire
instance (native slim + web fat) gains per-endpoint `world_hw_a/b`, and both
wire shaders interpolate `mix(hw_a, hw_b, which_end)` so each segment's band
narrows/widens linearly — depth-tested, and the linetype dash still applies.
Alignment (widths must match the tessellated points) is solved by building
both together: a new `TruckObject::TaperedLines(points, widths)`, filled by
`common::tapered_band_points` (straight = endpoints, arc = 16 samples with the
width lerped along it). to_truck routes a varying-width polyline here —
LwPolyline before the PLINEGEN split, Polyline2D at its return — while a
uniform-width one keeps the constant-band Contour path unchanged. `world_width`
stays the widest edge as the PDF-stroke + zoom-floor fallback.
Blocks fall back to the constant-widest band (`LocalWire` carries no
taper_widths).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A wide polyline's flat band was drawn by the hatch pipeline: a solid fill
with only draw-order depth. It ignored the entity linetype (the width
"crushed" the dash pattern) and showed through 3-D geometry in a tilted view.
Draw the band from the polyline's own centre-line wire instead. A new
`WireModel.world_width` (drawing units) makes the wire vertex shader expand
the quad by `world_half_width / world_per_pixel` (clamped to a hairline), so
the band is sized in world units and tracks zoom. Because it is now a wire:
- the GPU dash shader dashes the band by arc-length — the linetype applies;
- it is depth-tested at the polyline elevation — no draw-order see-through;
- no CPU band triangles or per-dash geometry are generated.
`world_half_width` rides in `WireConst` (native, replacing a pad) and the web
`WireInstance` (loc 14); both wire shaders branch on it. The tessellator sets
`world_width` from the LwPolyline width on the Contour / SegmentedLines arms,
and the model-space fill loop no longer emits a hatch band for any LwPolyline.
Pick still uses the band polygon (`pick_tris`).
Blocks keep the hatch band: `LocalWire` carries no `world_width` (finalize
zeroes it), so a scaled insert still scales its band correctly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rectangular paper viewports now go through the same GPU stencil clip as
non-rectangular ones: viewport_clip_boundary_ndc returns the four rect
corners (a full-target mask) instead of nothing, so the render path has
no rect/non-rect special case — every content viewport carries a clip
boundary and draws with stencil reference 0xFF.
With a single clip path, the old per-wire / -image / -wipeout scissor
mechanism is fully dead and removed: compute_*_scissors, the
*_pixel_scissors buffers, the project_scissor helper, the per-pass
set_scissor_rect blocks, the vp_scissor field threaded through
WireModel / ImageModel / HatchModel and their GPU mirrors (WireGpu /
ImageGpu / WipeoutGpu / HatchWebGpu), from_run's scissor parameter, the
scissor grouping in build_wire_buffers, and the is_arena_eligible gate.
Net -245 lines across 54 files.
Rectangular viewports still clip exactly via their own render rectangle;
the stencil mask is a redundant full-target no-op there, kept only so
the clip code path stays uniform. XCLIP keeps its CPU wire clip.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
MLINE dashed elements were fully CPU-expanded (apply_along) — one wire per
dash — to keep the parallel lines in phase. Move pure-dash elements onto the
GPU dash shader and end-align every element to the multiline's centre-line so
they share one interior phase.
- wire_distances now accumulates arc-length from the double-single (high+low)
points, so the dash phase stays precise at UTM coordinates. f32-high-only
deltas quantised ~0.1 there and drifted the parallel lines apart — the
original reason MLINE dashes were CPU-side.
- New WireModel.dash_align_end carries a shared begin/end dash length. The
shader uses it as align_end for every parallel element (the interior phase
depends on align_end, not the wire's own length) while align_total stays the
element's own length, so each still ends on a dash at its own endpoint.
- MLINE tessellation derives that length once from the centre-line, routes
pure dash/space/dot elements to the GPU pattern and text-bearing elements to
apply_along.
- apply_along gained an "A"-type mode (shared reference length): a solid
begin/end dash and a phased interior, so its dashes — and embedded text —
line up with the GPU-dashed sibling elements instead of tiling from start.
Pure-dash MLINEs now cost one WireModel + pattern per element instead of N
segments; single (non-MLINE) entities keep the from-start tiling unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
DET pipe walls now carry the stroke pattern (dash_from_start), which
tiles from each wall's own start vertex and skips the A-type end
alignment that was force-lighting a solid dash under the end symbol.
Two rendering fixes on top:
- Dedup the host polyline's vertices before offsetting the walls. Some
DGN pipe polylines carry trailing duplicate end vertices; a zero-length
segment gave offset_xy a degenerate normal, leaving that vertex
un-offset and folding the wall down to the centre line. The two walls'
fold segments met at the centre and drew a spurious line linking them.
- Re-key the symbol (end circle / half-ellipse) wires to the host handle
so the whole pipe picks and selects as one entity instead of the
symbols acting as separate phantom entities.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On a geometry edit the resident set gets a fresh content id, so the batched wire
path (build_wire_buffers) re-emits and re-uploads EVERY wire instance — the whole
buffer even when one line moved. Behind the OCS_WIRE_GPU_PATCH flag (default off,
native only), add a persistent per-entity instance arena: one long-lived
VERTEX|COPY_DST instance buffer + its WireConst storage, with a per-entity slab
table. An in-place edit (move / rotate / scale / colour — same segment count)
becomes a queue.write_buffer of just that entity`s slab.
Scope is deliberately IN-PLACE ONLY. draw_depth_map normalises each entity`s
draw-order z-bias by the block`s entity count, so an Add or Remove re-scales EVERY
entity`s z-bias — which a per-slab patch can`t honour — and a segment-count change
relocates the entity. patch() returns false for any of those (and for mesh-edge /
scissored sets), and the caller falls back to the identical full batched rebuild,
so correctness never rides on the fast path. An adversarial review of an earlier
append/tombstone design surfaced exactly these draw-order hazards; restricting to
in-place removes them while keeping the common interactive-edit win.
The Scene hands the changed handles to the arena via model_wire_gpu_patch,
recorded by try_resident_patch for the Model block and matched by content id in
prepare(); scissored paper viewports and mixed 2D/3D sets keep the batched path.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Dashed/dash-dot lines could end in a gap ("empty space at the end"),
especially inside paperspace viewports. The dash pattern was only centred
on each wire, which makes the two ends symmetric but does not force them
onto a solid dash.
Implement AutoCAD "A" (aligned) linetypes: the line begins AND ends on a
solid dash, interior dashes stay nominal, and the two end dashes stretch
symmetrically to absorb the remainder (so parallel lines share an
identical interior). wire_distances computes the end-dash length D and
total length L for a dash-first pattern (D = (L - k*P + a) / 2); the GPU
in_dash walk force-lights the two end regions and phases the interior to
resume right after the start dash. Lines shorter than one pattern period
draw solid.
Only alternating dash-first patterns are aligned; gap/dot-first,
consecutive-dash, NaN-break (plinegen=false) and solid lines keep the
prior behaviour. Native and web (wasm) shaders kept in sync.
Refs #13.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Mesh (PolyfaceMesh / PolygonMesh) outline edges are drawn through the wire
pipeline, whose shader pulls each wire toward the camera by
draw_depth * DRAW_ORDER_BIAS (0.001) for 2-D draw-order layering. Every
entity gets a draw_depth in [-1, 1], so a mesh's edges were pulled forward
too — and when the mesh's depth extent projects to less than the bias (a
small or distant mesh), its hidden back-face edges peeked through its own
shaded fill. Solid feature edges use the mesh shader, which has no such
bias, which is why only meshes showed it and why adjusting the fill's depth
bias had no effect.
3-D mesh outline edges are real geometry occluded by true depth, so force
their draw_depth to 0 (they are already grouped as `mesh_edge` in
upload_wires). They now occlude exactly like solid edges.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Remove the unused `crate::par::prelude` glob import in wire_gpu.rs and
the never-called `viewport_at_paper_point` method; fold its doc into the
screen-space twin that is actually used so no intra-doc link dangles.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Every wire segment is one instance, and each carried the wire's color,
line-weight, dash pattern and draw-depth (~44 B) inline — re-fetched per
segment even though they're constant along the wire. On native, move
those into a per-wire `WireConst` storage buffer (group 1) indexed by a
`wire_id` on the instance. The instance drops from 104 B to 60 B (~42 %)
across every entity's wires, and the shared constants are fetched once
per wire (cached) instead of once per segment — less vertex-attribute
bandwidth and less GPU upload on rebuilds.
WebGL2 has no vertex-stage storage buffers, so the wasm build keeps the
original self-contained fat instance and wire.wgsl; native uses the new
slim instance, wire_indexed.wgsl, and a group-1 bind group built per
batch in WireGpu::from_run / from_batch. Visual output is unchanged
(colors are now full f32 rather than 8-bit, if anything more precise).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Vertices now carry the low residual alongside the high half all the way
to the GPU: tessellation splits the offset-relative f64 coordinate into
(high, low) f32 (TruckTessResult::Lines/Point, the Text path, Lines /
SegmentedLines, and the block-cache instance transform), WireModel and
the block batch carry a paired points_low buffer, WireInstance gains
pos_a_low/pos_b_low (locations 10/11), and wire.wgsl reconstructs
rel = (pos_high - eye_high) + (pos_low - eye_low).
This is the piece the camera-only RTE was missing: at UTM-scale
coordinates the single-f32 vertex was quantized to ~0.5 m, so it snapped
each time the eye crossed an f32 boundary on pan. With the low half the
position is exact and the cross-drawing-paste jitter is gone.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Gate target-specific code so neither build warns about the other's items:
native-only helpers (view, window icon, file dialogs, file association, plugin
host/guard/registry, self-update) are cfg'd to non-wasm; the wasm-only font
fallback (script_of / cjk_lang / request) is cfg'd to wasm. Drop an unused mut,
allow the few genuinely-conditional dead items (mesh_click_hit, GPU layout
helpers, parity stubs), and gate the embedded PlotSettings-unrelated dead
constants. Both targets now build clean (only the unavoidable nom/quick-xml
future-incompat dependency notice remains).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
wasm32 has no threads, so rayon's global pool would panic at runtime.
Add `crate::par::prelude`: native re-exports `rayon::prelude`, wasm
provides sequential `std`-iterator equivalents (par_iter / into_par_iter
/ flat_map_iter over slices, Vec, HashMap, HashSet). All call sites now
`use crate::par::prelude::*;` instead of rayon directly.
- wire_gpu's rayon `reduce` (2-arg, no std equivalent) is cfg-split to
a sequential flat_map concat on wasm.
- rayon moves to a non-wasm target dependency.
Native still runs in parallel; all 75 tests pass. wasm bin compiles.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Move the 30 flat modules under src/scene/ into six folders by role:
convert (entity→geometry), text (fonts/shaping), model (render models),
pick (hit-test/selection/grips/xclip), view (camera/transform/render),
cache (block + property caches). pipeline/ is unchanged.
Full path migration — all `scene::X` references become
`scene::<group>::X`, no compatibility re-exports. render.rs items that
were `pub(super)` (parent was scene) become `pub(in crate::scene)` to
keep their original scene-wide visibility.
No behavior change; build and tests are green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Follow-up to 3.3. The selection xray pass still built one buffer + one
draw call per selected wire, so select-all over a large drawing issued a
draw per entity. That pass applies neither scissor nor mesh-edge skip, so
all selected wires merge into one order-preserving run via from_run.
Removes the now-dead WireGpu::new / build per-wire constructors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 3.3. upload_wires built one GPU instance buffer + one draw call per
WireModel — tens of thousands on a large drawing, a major per-frame CPU /
driver cost. Merge maximal runs of consecutive wires that share scissor +
mesh-edge state into one concatenated instance buffer each (WireGpu::
from_run); the existing draw loop then issues a single draw per run. A 2D
model (uniform None scissor, no mesh edges) collapses to one buffer + one
draw.
Runs are consecutive, never globally regrouped, so the sorted wire order
is preserved exactly — depth-biased overlap and alpha blending both
resolve in submission order, so the image is unchanged. from_run
guarantees instance order (index-ordered collect + sequential flatten).
The WireInstance layout, shader, scissor projection and draw loop are
untouched; only buffer packing changed. Selection overlay stays per-wire.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Phase 4.1. Handle / u64 keys dominate the hot maps (block_defn,
hatch / image / mesh caches, draw-depth map, viewport wire caches);
the default SipHash is wasted work on integer keys. Alias the std
types to rustc_hash::FxHashMap / FxHashSet crate-wide so every map and
set — including the cross-module draw-depth map shared between scene
and the GPU pipelines — uses the faster hasher consistently.
Mechanical: no behavioural change. Constructors moved from ::new() to
::default() (Fx variants carry no RandomState). Build + 13 tests green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Karim Jerbi