* claim: #4109
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Sketch: a preference to turn off scaling the sketch to its first dimension (#4109)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Test the first-dimension preference at small, typical and large values
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Engine: the scale_sketch doc names the Sketch preference
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Translate the first-dimension preference into the complete catalogs (ja, zh)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Add the Arabic, Hebrew and Vietnamese texts for the first-dimension preference
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
Add saved optical models, reciprocal rough interface transport, hard-light
connections, and bounded photon caustics for surfaces and selected media.
Keep inactive Glass sampling behavior and document admission and model limits.
Expose native controls and provide original CPU/GPU diagnostic captures.
* Orbit drags keep turning at the edge of the view
An orbit drag that reached the edge of the screen stopped turning: the cursor
stops there, so the pointer reports no more motion. Now, when an orbit drag
reaches the edge of the view, the cursor moves to the opposite edge and the
frame that reports the jump turns the view only by the hand's motion, so the
drag can go round the model as far as the user likes. Synthetic drags
(control channel, scripts) never move the real cursor. Where the cursor can't
be moved (Wayland, the browser) nothing changes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Orbit cursor wrap: a Navigation preference keeps the old behaviour
Wrapping the cursor during an orbit changes what an orbit drag does out of the
box, so Preferences › Navigation gets "Wrap the cursor while orbiting" (on by
default, `orbit_cursor_wrap` in ui.prefs). Off, wrap_cursor never moves the
cursor and the drag stops at the edge of the screen as before. The new strings
are a pending locale file, with Arabic, Hebrew, Japanese, Vietnamese and Chinese
for the complete catalogs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Use the current egui id calls
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: rootarchitect <7781866+rootarchitect@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* Point clouds: remember the source file and reload it (#8818)
A cloud keeps the absolute path it was inserted from. pointcloud.reload
reads it again, keeping the cloud's name, placement, units and colouring,
and the filters on properties the file still has; the Edit Point Cloud card
has a Reload from file button.
The file is read beside the window through the file loader, with progress
and Cancel, as an insert is. A thinned cloud stays thinned to as many points,
which reads an unchanged file as the same points; any other is read as a new
insert would be, so a file that grew comes in whole.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a reload asks before it reaches a network location (#8818)
A design names the cloud's file, and a design may come from anyone. On
Windows, only asking about `\\host\share\x` connects to that computer and
may send it the user's sign-in. So `pointcloud.reload` now reads through
`vfs::read_design_path`, as linked designs do, and Reload from file asks
with the same prompt as Update Links.
The button's label uses its stable message id. The strings gain Korean,
which a new label needs. `pointcloud.reload` gets its line in
tests/kernel-cases/EXEMPT, which the "Every command has a case" job needs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a reload reads under the limit the cloud was read under (#8818)
The limit was inferred from the counts, as if a cloud with fewer points than its
file had been thinned. A point that cannot be read makes the same difference:
such a PLY read as half its points on an unchanged file, since a limit one below
the file's count is a stride of two, and a file that grew was cut to the old count.
Keep the limit itself: a read is a function of the file and the limit.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a reload drops a value filter when the file stops naming the values (#8818)
A set of values filters a label, and `pointcloud.edit` refuses one on a property with no names.
A reload kept such a filter while the property stayed, so a file that dropped its names comment
left a cloud every later edit refused. One predicate now says what is refused and what a reload
drops.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a reload keeps its undo step and ends a play it moved (#8818)
A reload accepted with the Edit Point Cloud card open left the card's undo
boundary before it, so the card's next edit folded the reload into itself and
one Undo took back both. The boundary now moves past the reload.
A play told its filter by place alone. When a filter before it went, by a
reload, an agent's edit or an undo, the next filter took that place and the
play's range, on screen and written when the play ended. A play now names its
filter's property and ends, unwritten, when that place holds another.
`engine.batch` and the control hook compared command ids exactly, while the
engine takes any case: `POINTCLOUD.RELOAD` passed the refusal and read the file
on the window thread. Both now resolve the id as the engine does, for every
command that loads a file. The batch row of the protocol said the refusal held
only with the agent cursor shown; it always holds.
No test was run locally, at the owner's request: compile, clippy and fmt only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a file load keeps the request's own param_check (#8818)
`file_load::control` takes a command the file loader runs before
`control::handle` does, so that command skipped the request's `param_check`:
`pointcloud.reload` accepted a word that is no level, a `strict` call with an
unknown parameter read the file anyway, and the reply did not say the level.
`doc.open`, the imports and the DXF export went the same way.
The hook now takes the level itself and checks the call before the file is
touched, as a command that switches designs does: the worker's session goes on
to be the design's, which keeps its own level. Both share the same check, and
the reply says the level and any warnings.
No test was run locally, at the owner's request: compile, clippy and fmt only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* Design tabs: double-click a tab to rename the design
A double-click on the selected design's tab opens a name field in the tab
with the name selected, so typing replaces it. Enter or a click anywhere
else keeps the new name; Escape keeps the old one. The tab grows with the
name while it is typed (up to the tab's usual 200 pt). The rename goes
through component.rename with component 0, so it is undoable, marks the
design unsaved and is what Save As offers next. A name left as it was or
left blank changes nothing, and names are cut to the 128 characters the
command takes.
On desktop the bar is the window's title bar, and a double-click on a tab
used to maximize the window. Double-clicking the bar's empty space still
does; the tabs now rename. The preference tab_double_click_renames
(Preferences > Interface, "Double-click a design tab to rename it") brings
the old behaviour back.
Changes a default: yes - a double-click on a design tab renames the
design, old behaviour (maximize the window) kept behind preference
tab_double_click_renames.
Strings in assets/locales/pending/design-tab-rename.json (en, ar, de, he,
ja, ko, vi, zh). Docs: the UI tour and the preference list in
control-protocol.md (handle tab:rename).
Tests in documents.rs: Enter renames and undo restores; a click elsewhere
keeps the name and the tab grows while typing; Escape, a blank name and
the same name change nothing; an over-long name is cut to fit; with the
preference off a double-click does not rename.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Fix the tab rename for the current egui and ignore the rename result
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* claim: #12270
* Test: run --out writes meshes at the --quality asked for (#12270)
* Command line: run --quality standard|fine|finest sets the mesh quality of --out exports (#12270)
file.export already took quality, but run --out passed the path alone, so a
fine STL needed a script. --quality is checked before anything runs and goes
to every --out export; the usage text and the getting-started guide say so.
* Command line: run refuses --quality given twice, and the docs say the mesh chord never goes under 1 µm (#12270)
---------
Co-authored-by: Ahmad Khan <ahmad.khan@shiritai.co>
* Render: restore saved shots, progressive images and exports
* Render: register existing live and coat command tests
* Render: isolate lens saves and retain bounded previews
Save only lens parameters on existing CPU/GPU jobs while keeping unrelated
draft settings dirty. Retain pinned checkpoints through preview and status,
and release host-only snapshots outside the status lock.
Use the admitted PNG bound for registered-executor checkpoint slots. Preserve
executor work credits, final publication and regrade ownership. Add regression
coverage and align help, translations and capability documentation.
* Render: preserve live feedback and address review regressions
* Preserve saved cameras when applying Render library edits
Keep the saved camera when lighting or maps are applied to an existing shot. Fit and centre the measured Render host using the painted picture. Carry forward the upstream CPU and GPU preset and ZoomTool contracts.
Document::placements scanned every occurrence for each component it visited,
so its cost was placements times occurrences, on every world placement. The
new OccurrenceTree (crates/doc/src/occurrence_tree.rs) lists each component's
children once, in one pass and in document order; placements walks it with
the same order, caps and stop behaviour, through placements_with, which takes
each occurrence's transform from the caller. inside_component reads the same
index instead of scanning per component.
Tests compare the walk with the scanning one on 13 shapes and on hostile
generated documents, at its caps, in document order and with a given
transform. Release, 97,540 placements: about 120 ms -> 15.5 ms, the same in
either run order (docs/perf.md).
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* Appearance: theme picker with carousel, grid, filters and Day/Night slots
* Appearance: theme picker arrows move a screenful of themes
* Theme picker: translations, token metrics, wheel helper in navigation, reveal tolerance
* Theme picker: keep both new Appearance preferences after the rebase
* Appearance: adapt theme picker to egui 0.37
* claim: #2398
* Extrude: the distance arrow stops at a parallel face and outlines it (#2398)
While Extrude's distance arrow is dragged, a flat face parallel to the
profile and lying over it takes the distance exactly once the tip comes
within 8 screen points of its plane, and the face is outlined as a hovered
face is. A cut dragged through a block so stops at its far side. Faces
beside the profile don't catch the arrow; Alt/Ctrl (free drag) turns it off.
The distance stays a number in the existing parameter.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Extrude face stop: clippy (test-only current, +=)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Extrude face stops: a preference, on by default, keeps the old round steps (#2398)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Extrude face stops: follow world_state returning a Result (#2398)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Extrude face stops: import Vec/Plane from solidcore_record after the layer split
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* Open a design from a newer SolidCraft with a warning instead of refusing it
A newer format number or an unknown change no longer blocks opening. What
this version can read opens; a feature it can't read stays on the timeline
with an error, and fields, sections and container parts it doesn't know are
named in the reply and written back by Save, under the file's own number.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Saving a newer design from the CLI, and editing a feature this version can't read
`run --save` writes a design holding a feature only a newer build can build;
`--out` still stops, since it would export the part without that feature.
Edit Feature on such a feature says why instead of opening the palette.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Translate the newer-design notice for the complete catalogs; clippy
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #7273
* Viewport: drag a component as a GPU layer of its own
Each step of a Move triad drag placed the component anew, and the
viewport then rebuilt the model scene and uploaded it again. The dragged
bodies are now drawn from the scene built where the drag found them,
moved on the GPU by the drag's motion; the rest of the model stays as it
was. In the viewport a step meshes, builds and uploads no geometry;
only the transform (and the ground shadow) changes.
- gpu: Frame::moving draws a second scene with its own model matrix
(U.model, applied in every vertex entry, instanced parts included);
blocks, evaluated chunks and instanced triangles pass between the
model scene and the layer instead of being uploaded again; glass is
sorted in the layer's frame; the AO prepass and see-through pass
cover it; the frame key includes the layer.
- doc: ModelBody::display_moved_from gives the rigid motion between two
worlds' copies of one solid.
- ui-egui: the model scene without the dragged bodies is built once;
dragged bodies are matched by name; the step's body boxes are the
start boxes moved; selected or hovered dragged bodies light where they
stand. A drag back where it began keeps its layer. A section, an
analysis, continuity seams or no GPU draw the drag as before;
capturing the move makes the scene whole again.
* feat(ui): a setting to keep the highlight under the pointer while Move drags
A Move triad drag drawn as a GPU layer picks nothing under the pointer, which keeps a step from
meshing the dragged bodies where they stand and from looking up the Browser row a highlight
follows. Preferences > Tools > Assembly Move > "Highlight what is under the pointer while dragging
a component" (`move_drag_highlights`, off by default) picks and highlights as before.
Also: the drag layer test takes `as_chunks` (clippy 1.99's `chunks_exact_to_as_chunks`).
* fix(ui): say when the Move drag highlight setting applies
The label names Move's arrows, and the tooltip says that with a section or an analysis shown the drag highlights all the same, and that a part dragged along its joints never does.
* claim: tap marking records once in the model cache
* Model cache: write each tap marking record once and share it between states
Problem: every state of a saved model wrote the cosmetic tap markings of all the features
before it, so the records grew with the square of the features. A design of 400 marked
cylinders wrote 34 of its 37 MB manifest that way: past the 32 MiB limit the save thinned the
cache (590 of 806 states left out, opening evaluates them again) and wrote the manifest
several times while choosing what to keep.
Solution: the records keep a list of markings, each written once (keyed by its text, as the
sheet bodies are); a state names its markings as runs of places, like its bodies and
sketches. A marking at the same place in the state written just before, equal to the bit, is
found without writing it as text. A thinned model keeps the whole list and counts it once;
the reader counts the markings with what a state holds. The cache format is now /6: an older
cache is a miss and the design is evaluated as before. The design document is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Saved model: count each tap marking by its text against the held limit
A state clones every marking it lists, so a run repeating one large marking could hold far more than MAX_HELD items of memory.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Markus Kaltenbrunner <kaltenbrunnermarkus@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* Point clouds: a PLY file's edges and faces as beads and triangles (#5151)
A PLY file's `edge` and `face` elements become segments and triangles
between the cloud's points (a face as a fan), each with its own properties;
a file with them is read whole, since thinning would cut them. A point's
numbers vary along the segments and across the triangles that meet there,
an element's are constant on it.
Segments with `width` and `height` are drawn as beads of that stadium
cross-section, meeting mitred where one runs into the next, from one
indexed template; without them, as lines. Points no segment or triangle
ends on stay dots. A range filter cuts an element where its value crosses
an end, on the GPU and in the tracks alike: each element weighs its
measure (a dot 1, a segment its length, a triangle its area, each kind
over its own total) and spreads it exactly over the values it covers, a
segment's evenly and a triangle's as the tent its corners make.
Columns saved without the element they are on are on the points, so
designs saved with point properties alone still open.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: past max_points a PLY is thinned again; the near plane cuts elements exactly
A PLY with edges or faces and more vertices than `max_points` was refused, where it used to be
thinned. It is thinned again: thinning would cut its edges and faces, so they are left out, with
a warning, and raising `max_points` keeps them.
A segment or triangle crossing the near plane moved its corners behind the eye to a made-up
place, which pulled a wedge across the view. A triangle now keeps its true clip-space corners,
which the rasteriser clips exactly. A bead is cut where its axis crosses the plane, a crossing
exact because clip-space coordinates are affine along a line, and its numbers are exact there.
A ramp accepts MAX_BINS bands but an element was split into 4097, so a ramp finer than that lost
measure: half of a segment at 8192. Only the whole numbers have more bands than MAX_BINS.
`xtask mesh-decisions` reads any field named `triangles` as a mesh; a cloud's own complex holds
none, so its three modules are skipped with that reason.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a ramp's measure is whole past both its ends; a column on the points says nothing
A ramp of MAX_BINS bands has one band below it and one above it, so an element running past both
ends crosses MAX_BINS + 1 bands past its first. The limit was one short, and such an element lost
what lay above the ramp.
A column on the points now leaves `on` out when saved, so a design holding only points keeps the
form it had, as the field's comment and the pending format note already say.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: a set filter keeps part of a bead whose ends differ; an unreadable point takes only what ends on it
The set filter's whole-element test read the first corner alone, so a bead whose
first end was not kept vanished, while the range tracks counted the part kept. It
now drops a bead only when one value at every corner is not kept; otherwise the
fragments decide, as for triangles.
A PLY with edges or faces and one unreadable vertex was refused whole. It now
keeps the subcomplex the readable points span, renumbered, with its numbers, and
warns about what it left out. A file with numbers on its points gains the same.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Point clouds: build the triangle corner weights without a dynamic vector write
The Windows D3D12 shader compiler rejected w[vi] = 1.0 (error X3500).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* Book: sketching step by step, with three clips and a numbered screenshot
The sketching page now opens with a first sketch in four steps: start a sketch on the Top
plane, draw a 60 x 40 rectangle and a 12 mm circle with typed values, dimension the circle
until the sketch is fully constrained, and finish. Each step has a short clip, and the last a
screenshot numbering the Sketch tools, a constraint, a dimension, the sketch's state and
Finish Sketch.
doc_shots.py gains what these needed: a setup that runs before recording, so a clip starts
where the last one ended; sketch framing; Enter sent as two input frames; and clips that keep
their pauses (hovering a plane).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: the sketching page's general notes get a heading after Step by step
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: sketching adds Offset and Trim; the sketching figures again on 0.29.0
Two more clips after the sketching walkthrough: Offset draws the rectangle's outline 4 mm inside
it with one click (Chain), and Trim cuts a notch where a circle crosses the top side. All the
sketching figures are made again on SolidCraft 0.29.0, with the clips keeping a moment on each
result.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: solid features step by step, with four clips
The solid features page now opens with the sketch from the sketching page made into an open
box in four steps, each with a short clip: extrude it 12 mm, shell it with the top open,
round its outer top edges, then make it 20 mm tall from the timeline while the shell and the
fillet rebuild. The panel table and what follows get a heading of their own.
doc_shots.py: the feature shots share their steps with the setups that lead up to the next
shot, and fail when a feature on the timeline carries a warning or an error, so no figure
shows one. Shell comes before Fillet for that reason: shelling through the rounds needed the
kernel's boolean retry. trim_idle.py takes --hold, the seconds of each idle stretch to keep
(default 0.1 as before); the doc clips keep 0.8 s, so a reader sees each result. The app gets
up to three minutes to answer on a busy machine.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: features add Hole and Circular Pattern; the Parameters dialog, numbered
The features walkthrough gains two steps before the timeline edit: a 4 mm hole on the box's
floor, then Circular Pattern repeating it four times around the tube. The edit at the end now
rebuilds all of it. The parameters page gets a screenshot of the Parameters dialog numbering
apply as you go, user parameters, model parameters by feature, an expression and its value.
The features clips are made again on SolidCraft 0.29.0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: components and joints step by step, with three clips
The components and joints page now opens with a two-leaf hinge in three steps, each with a
short clip: the two bodies made into components from the Bodies folder's menu (the first one
grounded), a revolute joint on their barrel ends with Flip turning Leaf B the right way round,
and a drag that swings Leaf B about the hinge while grounded Leaf A stays put. What follows
gets a heading of its own.
doc_shots.py builds the leaves from commands, with Leaf B lifted and set back so the joint
visibly brings it into place.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: assemblies add Drive Joints; the assembly clips again on 0.29.0
A fourth step drives the hinge with Drive Joints: the slider swings Leaf B while a ghost shows
where it started, and OK keeps the new angle. The assembly clips are made again on SolidCraft
0.29.0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Book: getting started, sheet metal and import/export get pictures
Getting started gains a section on the start page: a numbered screenshot (new design, open,
import, recent, samples, the File menu) and a clip opening a sample from it. Sheet metal opens
with a tray in three steps, each with a clip: a base flange from a sketched rectangle, edge
flanges from two top edges, and the flat pattern with its bends. Import and export opens with
the File menu, numbered.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #11271
* Polygon By side: Length and Sides boxes, the typed length a side dimension (#11271)
The edge polygon tool showed no value boxes after its first click. It now
shows Length (the side from the first click) and Sides, as the inscribed
and circumscribed modes show Radius and Sides. A typed length places the
second corner and becomes a Length dimension on the first side:
sketch.draw.dimensioned takes a `length` role for sketch.polygon.edge,
an alternative to its radius (as_drawn still adds the radius when no
length is given; both together are refused).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Polygon By side test: read the design's own state so it compiles against main (#11271)
Main's Session::world_state() now returns a Result, so the new test failed
Quick compile on the merge. The sketch is in no component, so the design's
own state shows the same sketch on both the old and new bases.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Polygon By side test: take Vec2 from solidcore_record, where main moved it (#11271)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: thocsmith <193096300+thocsmith@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* Ribbon: an option to fold it to its tab row (#10112)
Preferences > Layout > "Show only the ribbon's tabs" (off by default), the
Ribbon Tabs Only command (view.ribbon_tabs_only), and its settings-profile key.
Rebuilt on main's redesigned Layout page.
* Ribbon Tabs Only: switching tabs with a panel menu open no longer keeps the overlay open (#10112)
A tab click now forgets the open panel drop-down of the previous tab, so the pointer leaving or Escape still closes the overlay.
* Ribbon Tabs Only: count its settings-profile key in the profile membership test (#10112)
* MCP: compact inspect_design and tool JSON, off by default (--compact / SOLIDCRAFT_MCP_COMPACT)
Split out of #2901 per review: the smaller inspect_design changed the default
for everyone, so it is now an MCP server option that is off unless asked for.
- solidcraft-cli mcp --compact, or SOLIDCRAFT_MCP_COMPACT=1 (true/yes/on;
0/false/no/off/empty leave it off; anything else is an error at start-up).
solidcraft_mcp::Options::resolve is the one place that decides, so an app
Preferences setting can feed it later.
- With it on, inspect_design summarises sketches (id, name, plane, counts of
points, curves, constraints and profiles, dof, status), returns full contents
only for the sketches named in `sketches` (names/ids, true, false or
"summary"), takes a `bodies` filter and measures only those; unknown names
list the existing ones. Tool JSON is compact rather than pretty-printed.
- Off, tools/list, argument checks, inspect_design and tool JSON are exactly
as before.
- document.inspect gains an opt-in `sketch_counts` (plane, points,
constraints); without it its reply is unchanged.
Tests: both modes in the MCP suite, Options parsing, and the CLI flag and
variable end to end.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* MCP compact output: a Preferences setting too (AI › Agents through MCP)
The desktop app's Preferences › AI gets "Compact output" under a new "Agents
through MCP" section, off by default and shown with the chat off too. The app
reports it on the control channel as `mcp.options` ({"compact": bool}); the
agent chat lets its own MCP server read it.
A bridged MCP server (`solidcraft-cli mcp --connect`) asks the app when an
agent first lists or calls a tool and keeps the answer for the session, so the
tools it listed keep their shape. An app not running yet, or busy past 2 s, is
asked again on the next call; an app from before the setting counts as off.
`--compact` and SOLIDCRAFT_MCP_COMPACT=1 still turn it on whatever the app says.
Tests: the server follows the app's answer (waits for it, keeps it, the flag
still wins); the checkbox reaches `mcp.options` and Restore Defaults clears it;
the chat gate allows `mcp.options`. docs/mcp.md and docs/control-protocol.md.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* MCP compact output: review fixes (mcp.options never blocks, headless test, locales, prefs doc)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* MCP compact output: the setting is a section of the AI page; the bridged test answers mcp.options
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* MCP compact output: the preference test sets it through ui.prefs; rustfmt
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* MCP compact output: look for a newly started app before asking its options; counts from sketch.inspect for an older app
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #10745
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Control channel: ui.menu refuses missing targets, reports the History item menu, and tells the joint menu items apart (#10745)
- ui.menu looks a target up and refuses one the design doesn't have ("no body `NoSuchBody`").
- The History bar's feature menu is the shared context menu (Target::Timeline), so ui.menu
reports it, ui.menuPick runs its items and {"type":"timeline","feature":id} opens it.
Its Show All toggle moves to TreeState so the menu can list Delete after the marker.
- Edit Joint Limits is ui.editJointLimits and opens Edit Joint with Limits expanded.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Icon audit: open the section and canvas menus on a real section and canvas; a hidden section has a menu too
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ui.menu: check a group's or folder's scope and folder name; Delete after the marker gets its own id
A group or folder target is refused unless its scope is a row of its component (`c<id>` or a
placement of it) and the folder is bodies, sketches or construction. The History item menu's
delete of everything after the marker is ui.deleteAfterMarker, so ui.menuPick can tell it from
Delete (review on #10746).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ui.menu: a component's own c<id> scope is the design root's only; test that Delete after the marker deletes
The browser draws a part's folders and groups under its placements (o:<path>), so a c<id> scope
for any component but the root names a row that isn't there (review on #10746). The History test
now picks ui.deleteAfterMarker and checks the features are gone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ui.menu: a placement scope must be spelt as the browser keys it
o:012 parses to placement 12 but isn't the row's key (o:12), so Show/Hide would store a key no
row reads (review on #10746).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* History item menu: no Delete after the marker when nothing shown is after it
A timeline target for a feature the bar's component filter leaves out listed no features, so the
item did nothing (review on #10746).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ui.menu: a plane target must be construction geometry
The plane menu's Show/Hide keys the feature's name as an origin item, so any other feature got a
menu whose actions don't fit it (review on #10746).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ui.menu: accept the view navigator's and selection filter's menus, which main added
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #7641
* Sketch on a thickened surface follows the face (#7641)
Thicken dropped the kernel's face history, so Create Sketch refused
every face of the solid. Carry that history through, the same way
Shell and Trim do.
The regression covers the offset face and an end wall, and checks
that a boss on the offset face follows a thickness change and reopens.
* Claim #9736
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Assembly: a move not captured yet is an unsaved change (#9736)
A component moved without capturing its position left the design reading
as saved: file.save wrote the file without the move and said nothing, and
closing the tab didn't ask. Now:
- Session::is_dirty is true while a move is pending, so closing a tab or
quitting asks "save changes?".
- file.save and file.save_as refuse while a move is pending, naming the
component and the capture and revert commands.
- In the app, Save As (and Save As JSON, Save a copy, Download design)
asks the Capture Position question first, as Save already did; a tab
closing after "Save" waits for that answer, and Cancel keeps it open.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Close: the save-changes card waits behind the Capture Position question (#9736)
Saving a tab that closes asks the Capture Position question first; the
"save changes?" card stayed drawn over it and took a second answer. It
now hides and ignores answers until the question is answered. The design
container doc says a save stops for a move not captured.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Assembly: an exploded view shown is not a move to save (#9736)
explode.show displays a view through the pending moves, as an uncaptured
move does. Those placements last only while the view is shown, so they
don't make the design unsaved and don't stop a save
(Session::moved_uncaptured). A move on top of a shown view still does.
The relink test saved with a move pending; it captures it first now.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Assembly: the pending-move save test reads the world state as main now gives it
world_state() returns a Result since the kernel merge; the test unwraps it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Docs: add BoxInsertDesigner to the showcase
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Docs: attribution rows for the BoxInsertDesigner showcase recording
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: the component menu keeps its everyday actions and groups the rest in flyouts
A component's or placement's right-click menu had grown to about thirty entries in seven
blocks. Activate, Edit in Place, Open Linked Design, Move/Copy, Show/Hide, Isolate, the
clipboard, Delete, Rename, Appearance, Physical Material and Properties stay on the menu;
Ground, Capture and Revert Position, Unlock Contents and Turn Up Axis move to a Position
flyout, Relink Source or Extract to File and the exports to File, New Component, New Group
and Paste New to New, and Show All Bodies, Cycle Component Color, Expand All and Collapse
All to Display. context_menu::entries lists a menu with its flyouts, as ui.menuPick reads it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: the component menu nests its flyouts by topic, and flyouts no longer scroll
Position (Ground, Capture and Revert Position, Unlock Contents, Turn Up Axis), Component
(New Component, Rename, Appearance, Physical Material, Cycle Component Color, Relink Source
or Extract to File, Properties; Design on the design's row), Export (Export, Save As Mesh,
Send to Slicer) and Browser (New Group, Expand All, Collapse All) each open a flyout; Paste
New stays with the clipboard and Show All Bodies with Show/Hide. Flyouts open inside flyouts,
so Appearance keeps its quick choices inside Component; the keys walk in and out of them.
A flyout no longer scrolls or sits away from its row: every flyout has its own area, measured
whole, where a short one used to squeeze every later one into its height. A component's only
body is no longer offered as a component of its own on a folded Bodies folder's entries.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: the component menu starts with what a double-click does, and keeps the old order of its sections
A linked component's menu starts with Edit in Place and Open Linked Design when a double-click
edits it in place (double_click_components), and with Activate otherwise. Position, Component
and Export follow Move/Copy, then the clipboard, then Show/Hide and Isolate, with Browser last,
as the sections came before the flyouts.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: Show All Bodies is on a component's menu only while a body is hidden
It is listed while a body is hidden itself, in a hidden placement, or in a hidden Bodies folder or
group of bodies, anywhere in the design, since that is what it shows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: Relink Source and Extract to File sit right below Rename in the Component flyout
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: Placement and Organize flyouts, Turn Up Axis back as a flyout, and flyout fixes
Placement (Ground, Turn Up Axis, Capture and Revert Position, Unlock Contents) follows
Move/Copy above the first separator, Turn Up Axis opens its two turns in a flyout again, and
the Browser flyout is called Organize. The Component flyout has three blocks: New Component
with Relink Source or Extract to File, then Appearance, Physical Material and Cycle Component
Color, then Rename and Properties.
A flyout of a flyout that opened to the left opens to the left too instead of covering the
menu, Left back to the menu chooses the row that opened the flyout, ui.menuPick refuses a row
that only opens a flyout, and the icon audit checks every level. The guides and the control
protocol describe the new menu paths.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Browser: ui.menu gives a flyout's rows their placeholder flag too
ui.menu set placeholder only on the top-level rows, so a client reading
a flyout row's placeholder got nothing. The icon audit now reads the flag
of the flyout rows instead of resolving their icons itself.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Adds DeepSeek and Qwen presets to both endpoint protocols: OpenAI-compatible (api.deepseek.com, dashscope-intl compatible-mode/v1) and Anthropic-compatible (api.deepseek.com/anthropic, dashscope-intl apps/anthropic).
Co-authored-by: Claude Code <noreply@anthropic.com>
* claim: #2217
* Dialogs: drag a value's label sideways to change it (#2217)
A value row's label (expr_row, hint_row) now scrubs a literal value: about a
hundredth of its size per point of drag, on that step's grid, Shift ten times
finer, Ctrl/Cmd ten times coarser. The unit and decimal mark stay as typed, a
size never turns negative, and expressions and parameters are left alone. The
text changes as if typed, so previews follow and OK still makes one command.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Label drag: measure from the press point, cap values, keep scrub kit-private
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #8220
* Agent chat: Custom instructions in Preferences, added to the system note (#8220)
Preferences > AI gets a multi-line Custom instructions field, saved with
the other AI preferences (agent_instructions, kept by page resets like
other user-written text). The chat adds the text after its own system
note when a session starts, by the routes #9811 set up: ACP session/new
(the Claude adapter's _meta key, the standard systemPrompt field, the
Antigravity key) and the system message or field of every API protocol.
Connect and New chat set it, so an edit applies from the next chat and
never changes the one running. Blank text adds nothing; at most 4000
characters are kept, in the field, over ui.agent_chat and from a hand-
edited settings file. ui.agent_chat takes and reports `instructions`.
* Custom instructions: typing in the field stays in the field
The text area did not report itself to Preferences search as a key sink, so with
the field focused, search took the typed characters and moved focus away. The
field now notes the sink while it has focus, like the other preference fields,
and the #8220 test types into it.
---------
Co-authored-by: MateuszSychXpl <151755914+MateuszSychXpl@users.noreply.github.com>
* claim: #9236
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* GPU path tracer: don't rerun a setup that crashed in the driver (#9236)
A driver that takes the process down while it compiles the path tracer
(RADV on an RX 7600) is beyond error scopes and unwind guards. A running
setup now keeps gpu-path-setup.json (its adapter key) in the settings
folder, removed when the setup answers or the app closes normally. A
launch that finds it for the same adapter leaves the setup failed with an
explanation and a Try again button; CPU ray tracing stays available, and
a driver update changes the key and so tries again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* Appearance: a picture of your own behind the model
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Background picture: a label that fits its row
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #8768
* feat: Browser folders start closed and find a body only when asked, behind a preference
Preferences › Layout › Browser rows start closed (`browser_rows_start_closed`, off by
default) starts the Browser's folders closed, and placements too whatever "Components start
open" (#9274) says. Rows then open with their arrow, Expand All or Find in Browser, which opens
just the rows down to its body and closes again the ones the last find opened. A body picked in
the 3D view leaves the Browser as it is, and only the body Find in Browser found has the rows
above it marked, so a large assembly's placements are not named again after each edit.
With the preference off, folders start open, placements start as "Components start open"
says, and a pick in the 3D view opens the Browser down to its body, as on main. Find in Browser
looks the body's placement up once and opens its rows before drawing, instead of each row
testing the revealed path while it is drawn.
* Browser tests: read the row order from the Browser's content frame
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* claim: #9719
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Dynamics: rigid-body solver core with joints, gravity and springs (#9719)
A new L1 crate, solidcraft-dynamics (geom only): rigid bodies in maximal coordinates, joint constraint equations with analytic Jacobians for rigid, revolute, slider, cylindrical, planar and ball joints, gravity, linear spring-dampers and constant loads; the index-1 equations of motion solved through a rank-revealing Schur complement (repeated joint equations are solved, not refused); RK4 and Dormand-Prince 5(4) with position and velocity projection; joint values, rates, reactions and energies at every output sample; caps on bodies, joints, rows, steps and samples; Result errors only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* Dynamics: take Vec3 from the record crate after the layer split, refresh the lock file
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: bherbruck <brennen.herbruck@gmail.com>
* claim: #10220
* Web app: keep unsaved work in the browser instead of the leave-page box (#10220)
A new preference, off by default, on Preferences > General > Saving (web
app only): "Keep unsaved work in this browser instead of asking when the
tab closes". With it on, each open design with unsaved changes keeps a
copy in the browser's private storage (keep-<page>-<design>.sc/.json),
written once the design has been still for 0.8 s (at most 4 s after the
first change it lacks) and again on pagehide or when the page is hidden.
While every copy is current, beforeunload doesn't ask. A copy goes when
its design is saved, closed, or the preference is turned off.
The next visit offers the copies newer than their design's last save in
SolidCraft's own window, "Restore unsaved design: <name>, changed
<time>?", one window for several, with Restore, Open saved version, and
Discard. A page holds a Web Lock while it runs, so another open tab's
copies are never offered or removed.
With the preference off, the single autosave and the browser's prompt
work exactly as before.
* Web app: the restore window comes from the UI kit's recovery module (#10220)
The kit's text and divider helpers are private to the UI crate, so the
"Restore unsaved design?" window moves to recovery_ui
(browser_copies_window), beside the desktop recovery window, and the web
shell only maps its answer. Also keeps the window handle for the page
listeners instead of using the one moved into the app.
* Web app: restore window buttons sit in one row (#10220)
* Web app: keep unsaved work and still ask, unless told not to (#10220)
Keeping a copy in the browser no longer replaces the browser's leave-page
box by itself. A second option, shown once work is kept and on by default,
leaves the box in place; off, the box appears only while a copy isn't current.
* Web app: kept-copy review fixes (#10220)
A failed write of a kept copy waits 30 s before the next try, a recovery
copy is dropped only after the remembered saved file has opened, the
preference is documented under its current label, and the ui.prefs list
is sorted again.
* claim: #10223
* Quick-access bar: pin ribbon commands to the application bar (#10223)
A command's menu (a ribbon button, a section drop-down row or a command
search result) has Add to Quick Access: the command's own icon joins the
application bar after Save, Undo and Redo. A click starts it as its
ribbon button does, a drag along the bar reorders the buttons, and the
button's menu has Remove from Quick Access. Empty, and so not drawn, by
default; it holds up to 24 commands.
The list is kept with the toolbar's customisations (`quick_access` in
the toolbar's saved document), so it is saved, exported with the
keyboard shortcuts and carried by settings profiles.
* Quick access: a pinned family variant no longer replaces the ribbon choice
A quick-access click now starts the pinned command with start_pinned, so it
starts exactly that command and leaves the sketch family remembered choice
(the main ribbon button) unchanged.
github-actions[bot]