- Add 'i ' prefix to EntryKind::Info and coordinate set_step_prompt deduplication to satisfy WCAG 1.4.1 across all 22 themes
- Drop unused _desk parameter from crosshair_background and clean up its call sites
- Document AutoCAD-standard context-based crosshair drafting surface architecture in crosshair_background
- Add doc comments documenting deliberate alpha omission in canvas_is_light and backlog scope in is_wide
- Add canvas_is_light helper based on WCAG luminance (> 0.5)
- Ensure crosshair_background always returns sheet white in paper space
- Update resolve_selection_base_color to resolve user ACI -> light-canvas palette -> theme palette
- Add light_canvas_color with curated deep-green/deep-blue pairs for light surfaces
- Extract selection_fill_alpha with 1.35x contrast compensation on light surfaces
- Thread canvas_bg through draw_marquee and poly selection
- Note deliberate limitation for desk-spanning marquee edge case
- Add test coverage for paper space, ClassicDark, custom light bg, and fill alpha clamping
Everything measured so far accounts for about 1 ms of a ~120 ms pan
frame: update is 5-28 ms at worst and usually far less, prepare 0.45 ms,
encode 0.5 ms. Meanwhile the GPU idles at 0-8%, encoding no scene at all
leaves latency unchanged, major page faults stay flat at 17 across forty
seconds of panning, and freeing 6 GB of swap changes nothing either.
So the time is in code that has never been timed. `view()` runs between
the input handler returning and `prepare` being called, rebuilds the
whole widget tree every frame, and is the largest unmeasured span in
that gap.
Reports only when it exceeds 1 ms, so a healthy frame stays quiet.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Pan and even plain cursor movement over the canvas ran at ~8 fps on a
large drawing. Nothing rendered was responsible: the GPU idled at 0-8%,
encoding no scene at all changed nothing, and vertices, pixels and draw
calls could each be cut without moving the number. `view` was the frame
-- 110 ms of it against 3-6 ms on a small drawing -- and inside `view`
it was the status bar.
The status bar builds the selection-filter menu, which asks for the
distinct entity type names in the layout. That call cloned the layout's
entire handle list (1 166 357 handles, ~9 MB), then did a map lookup, a
String allocation and a BTreeSet insert for each one. About 107 ms of a
110 ms frame, to produce roughly twenty names, for a menu that is only
visible when opened.
Cache the result against geometry_epoch and the layout block, since it
only changes when entities are added or removed, and make the scan
itself borrow the handle list rather than clone it and collect &str,
allocating only for the names it returns.
The immediate-mode UI rebuilds every widget on every mouse move, which is
fine as long as nothing in that path is O(document). This was.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Model space appearance now follows the active UI theme instead of a
hardcoded dark background. Canvas mode is user-selectable (Match Theme /
Classic CAD Dark / Custom RGB), with matching controls for the paper
sheet and desk surround backgrounds and grid opacity.
Implements industry-standard CAD system variables, all settable via
command name, SETVAR, and the Options dialog:
- COLORTHEME (0/1 dark/light UI scheme, preserves custom palettes)
- COLORSCHEME (interactive named-theme selection)
- BACKGROUND (THEME/CLASSIC/DEFAULT/DESK presets, RGB triplets)
- SELECTIONAREA, SELECTIONAREAOPACITY (translucent marquee fill)
- WINDOWSAREACOLOR/WINDOWAREACOLOR, CROSSINGAREACOLOR (ACI or theme)
- SELECTIONEFFECT, SELECTIONEFFECTCOLOR (GPU highlight pass, driven by
selection_generation invalidation — no retessellation)
- SELECTIONPREVIEW (bit 0 = idle rollover, bit 1 = in-command hover)
- GRIPSIZE, GRIPCOLOR, GRIPHOT, GRIPHOVER (grip overlay styling)
Scene.selection was Rc<RefCell<SelectionState>> (31 fields, ~1.6KB
with poly_points) cloned deep per view_main per pane via
(*sel_ref).clone() at view/mod.rs:760. Replace with
Arc<RefCell<SelectionState>> (Arc::new(RefCell::new(...)) at
scene/mod.rs:1436,1868), SelectionCanvas holds Arc<RefCell> and
selection_overlay takes Arc<RefCell> (overlay.rs:365,415),
view_main does Arc::clone(&tab.scene.selection) (atomic bump)
instead of deep clone. sel_ref stays borrowed for field reads
(vp_size etc.).
Bench (release, n=5000): SelectionState::clone 39-68ns (64pts) /
11-22ns (0pts) vs Arc::clone 7ns — 5-9x on the hot path.
Tests:
- selection_is_arc_not_refcell: asserts type_name contains Arc
- bench_selection_state_clone / bench_selection_arc_clone / bench_selection_state_clone_empty (ignored, release)
Hold a single Ref<SelectionState> across the block; replace six redundant borrow().vp_size/last_move_pos calls with field reads on the Ref. Move the deep clone to the widget call site ((*sel_ref).clone()), preserving one clone per frame but ensuring the widget receives the same snapshot as the geometry reads.
Replace manual key entry with per-row capture cells: clicking + Add
inserts a draft row at the top with its key cell armed, and the next
key combination is recorded and normalized (Ctrl+Shift+S →
CTRL+SHIFT+S). While armed, a dedicated keyboard listener swallows all
input so recorded shortcuts do not execute in the drawing.
Draft rows are transient and confirmed only through an explicit accept
button, so the table never retains empty rows. Duplicate keys and
unrunnable commands are flagged in danger styling with a persistent
warning naming the conflicting command. Added reset-to-defaults with
confirmation, apply-and-exit, a shortcut count, and an unsaved-changes
guard on close.
Improve the selection-cycling list box so overlapping objects can be
identified at a glance:
- Add a color swatch per row showing the object's effective drawn color
- Show the layer name right-aligned next to the element type
- Size the popup to fit its content with consistent padding
Widen render_style_for_viewport visibility from scene-internal to
crate-wide so the view layer resolves the same color the renderer uses.
Persist camera, render state, and solid fallback geometry across saves. Generate previews from a clean live viewport and refresh recent-file thumbnails immediately.
The MText editor paints its own caret and is not a focusable widget, so
the command input keeps widget focus behind the dialog. It captures the
arrow itself, which routes the key through CommandLineArrowProbe instead
of ArrowKeyPressed — and that path only asked whether the command input
was focused, never whether an editor was open. Up and Down therefore
walked command history while the user was typing in MText.
Give the probe the same rule ArrowKeyPressed already applies: while the
editor is open the arrow moves its caret. The probe now carries the shift
state so a held Shift still extends the selection on this path.
Fixes#655
The ribbon kept its own \show_block_palette\ copy that the app had to
manually sync at every mutation point (the BLOCKPALETTE command and the
dock Close handler). Drop the copy and thread the app's authoritative
\show_block_palette\ into \Ribbon::view\ at render time, so the button
highlight can't drift from the panel's true visibility — those two
handlers are now mutate-points only and the ribbon view is the single
place that reads the state back.
The drag preview divided the edge height by the raw panel count on the target
side. Closed panels keep their stack slot, so a hidden panel (e.g. the block
palette tucked away while Properties stays open) split the ghost and the drop
index as if it were rendering — leaving a half-height preview on that edge.
Size slots from the panels actually rendered on the edge instead, via a shared
visibility helper used by both the drop-index math and the ghost/edge-tint
geometry.
The dock drag preview computed slot heights and drop indices from the full
window height, but docked panel columns only span the model viewport, so the
preview overlay and insertion point drifted below the actual dock edge. Use the
current viewport height so the ghost and drop index line up with the panels.
Clicking an editable property value field now selects its entire value instead
of only placing the caret, so typing immediately replaces it. The click is
observed through a global left-press subscription (PropPointerPressed); a focus
sweep reports which widget took focus and PropSyncActive maps it back to the
active property row, then issues a select-all operation only when focus landed
on a different field.
The active row stays highlighted while its text input holds keyboard focus and
clears when focus moves elsewhere.
Style the dock rail tab by its panel state: the expanded (active) panel's tab
gets a primary-colored highlight and border, and a tab whose panel is being
dragged gets a lighter primary tint, so the user can see which panel owns the
edge at a glance.
A paper layout painted its whole canvas in the page colour, so the sheet
had no visible edge — zooming out never reached one, because the surround
was the same colour as the page. Floating viewports then painted
themselves black over it.
Both come from the same idea, that a viewport's canvas is a colour it can
be handed. What lies behind a viewport depends on what the viewport is:
- The full-canvas sheet shows the desk. `paper_bg_color` is the page, and
the page is a bounded rectangle `paper_sheet_fill` already draws; using
it for the canvas too erased the boundary between them. The desk is now
its own colour, defined once in the scene and read by both the widget
and the renderer, so the two cannot drift apart.
- A floating viewport sits on that sheet and must paint no canvas at all.
Handing it a transparent one does not work: the background shader
returns an opaque colour whatever alpha it is given, so alpha zero came
out black. Skipping the pass is the only way to ask for see-through, so
`skip_background` now suppresses both the clear and the pass. It is part
of the frame signature too — without that, a cached frame would keep
showing the canvas the flag was just turned off for.
Shading and fog still blend toward what is behind a floating viewport,
which is the page, so the uniform keeps that colour even though nothing
paints it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The readout learned to write lengths as 5'-6 1/2" and angles as
N 45d0'0" E, but typed input was still a bare parse::<f64>() — a
coordinate could be displayed and not entered. Give the formatters
their inverses.
parse_length takes feet and inches separated by a dash, a space, or
nothing, with the closing quote optional, and whole-plus-fraction in
any of those forms: 5', 60", 5'-9", 5'9-1/2, 9-1/2, 1/2. A leading
minus is a sign; the one inside 5'-9" is a separator, since after feet
there is nothing left to subtract from.
parse_angle reads the mark that says which convention an angle is in —
g for grads, r for radians, d/'/" for degrees-minutes-seconds, compass
letters for a surveyor's bearing — and reads a bare number in whichever
convention the drawing is set to.
ANGBASE and ANGDIR now apply, but only where they mean something. Which
way a thing points is measured from the drawing's zero and runs the way
it says; how wide an arc opens is the same number whatever zero is
counted from. format_direction and parse_direction are the pair for
directions, and the polar readout and polar input both go through them,
so the bearing shown is the bearing accepted. format_angle stays for
sizes, where applying a base angle would corrupt the value.
The command prompts follow the same rule, each classified rather than
swept: OFFSET distance, FILLET radius, CHAMFER distances, ARRAY and
MINSERT spacing, TRACE width, HELIX, and EXTRUDE height read lengths;
the four dimension text-rotation prompts, ARRAY's total angle and
INSERT's rotation read angles. MLINE's scale and INSERT's scale factor
are left alone — a multiplier has no unit to be written in. INSERT read
both through one parse; they are now separate.
The typed variants also take a decimal comma, which a single-value
prompt has no other use for. parse_length itself does not: inside a
coordinate the comma separates the axes.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Hakan Seven