Commit Graph
100 Commits
Author SHA1 Message Date
jyvikoandClaude Opus 5.5 febcffefd6 verification: series-resistance heat flow across a tie (#358)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 08:27:01 -04:00
jyvikoandClaude Opus 5.5 836513ddfe solver: heat transfer across ties (#358)
Ties constrain temperatures with the displacement tie's mortar weights in
steady and transient heat and in coupled temperature-displacement steps; the
heat solver eliminates them with the prescribed temperatures through
constraints::Reduction. refuse_ties is gone.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 08:27:01 -04:00
jyvikoandClaude Opus 5.5 0ec1882697 solver: tie weights and constraint reduction independent of the structural mesh (#358)
The tie's mortar rows take any node numbering and degree-of-freedom kinds, and
Reduction takes a DofLayout, so heat transfer can reuse both.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 08:27:01 -04:00
jyvikoandClaude Opus 5.5 507e8e32d2 release: version 0.5.9
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 02:57:11 -04:00
jyvikoandClaude Opus 5.5 491b3f6212 engine: the heated-bracket guard checks its element count, not wall time
The 60 s limit failed on the Linux CI heavy job (106 s in the test profile) while the
mesh was within its band; the count is what #303's blow-up changes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 02:37:31 -04:00
jyvikoandClaude Opus 5.5 111a6f1a94 Mesh: a face edge split that encroaches a feature edge splits the feature edge
Delaunay refinement split a boundary face edge at its midpoint even when that
midpoint lay inside the diametral sphere of a feature edge beside it. Each such
split halves the face vertices' distance to the curve, so they piled up against
it (the heated bracket grew face vertices 0.016 mm from a 4 mm-sized edge), the
improvement schedule could not lift the tetrahedra there to its target, and
inserted vertex after vertex around them: 29257 elements and about two minutes
to open instead of 23641 and 6 s. The SolidCraft pin move (4c91f9b) added one
interior vertex to the bracket's end faces, which changed the CDT enough to
expose it.

Following Shewchuk (1998), an encroaching midpoint now splits the encroached
feature edge instead.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 02:17:00 -04:00
jyvikoandClaude Opus 5.5 43e2144090 Docs: the crank's one-datum workflow, its gallery images and the Decompose status
examples.md and mcp.md tell the one-datum workflow and its numbers; the
crank gallery captures and Start card thumbnails are taken again on this
build; ROADMAP marks the automatic blocks of extruded cross-sections done,
and features.md's Decompose entry says what is left (3D parts).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:50:49 -04:00
jyvikoandClaude Opus 5.5 8e45f2c275 Samples: the crank meshed from one datum, its half extruded in automatic blocks
crank-arm and crank-mirror now take one datum (Sym) and its partition:
mesh.hex.generate extrudes the half CrankArm/a, cutting its cross-section
into blocks itself, with no seeds, and mesh.mirror reflects it. 8 664
hex20, scaled Jacobian min 0.371; root sigma_xx -62.44/+62.44 MPa against
M/Z 62.6 (0.25 %), spindle reaction 1500 N. The seven partitions and
eleven exact seeds go.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:50:49 -04:00
jyvikoandClaude Opus 5.5 86ba1bda5c Release: sign SHA256SUMS.txt with the release key; docs/updates.md (#231)
The release job signs SHA256SUMS.txt into SHA256SUMS.txt.minisig with rsign2 when the
FIELDCRAFT_SIGNING_KEY secret exists, and checks the signature against the public key the app
is built with; without the secret the draft is unsigned, the run warns and the notes say the
app cannot install that release itself. docs/releasing.md says how the key is set up, kept
and rotated; docs/updates.md describes checking, verifying and installing, the web app's
Reload to update, and FIELDCRAFT_UPDATE_HOST for tests. privacy.md, control-protocol.md,
web.md and the roadmap say what changed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 93b2ae343c Web: offer Reload to update when a new deployment is out (#231)
Every half hour the page fetches its own address past the browser's cache; when it loads
another content-hashed build of the app than the running one, the status bar offers "Reload
to update" (the page's unsaved-changes question still guards the reload). Rust only, no
service worker: the browser build has no other update path and none of the desktop's
download code.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 7dd50a3ca4 App: run update jobs in the background and swap the release in after closing (#231)
The desktop app checks, downloads, verifies and stages on its own threads for the window, and
once the window has closed (autosave closed) swaps a staged release in and starts it. Staging
folders an earlier update left are removed at start-up. Automatic checks stay off under
--control automation and CI unless FIELDCRAFT_UPDATE_HOST names a test server.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 2e28a69c04 UI: update notice, Help › Check for Updates and Reinstall, Settings › Updates (#231)
The window checks at start-up at most once a day when automatic checks are on (Settings and
the first run's question), never while that question is open and never without a host, so the
browser and tests make no network call. A newer release shows a quiet notice in the status bar
("FieldCraft X is available — What's new · Install · Skip"); Install downloads and verifies
through the host, then asks to restart, saving or asking about unsaved changes first, and the
host swaps the release in after the window closes. A .deb/.rpm install offers the system's
installer and the command instead. Skip this version is kept in the preferences.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 69c1440cbe CLI: fieldcraft-cli update --check | --install (#231)
--check runs app.update.check and prints its JSON; --install downloads the latest release for
this install, verifies the checksum list's signature and the asset's checksum, and swaps it in
(or, for .deb/.rpm, prints the package manager command for the verified download). Tested
against a local release server through FIELDCRAFT_UPDATE_HOST.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 12487794a3 Engine: app.update.check reports whether a newer release is published (#231)
The same check as the app's Help › Check for Updates and fieldcraft-cli update --check: the
latest release, whether it is newer, its page, whether it is signed and the asset this install
would download. It only reports. Checking reaches the network, so only the desktop app and the
CLI switch it on; tests, scratch sessions and the browser build refuse without asking.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 8fe31545ea Update: check, verify and stage releases in a new fieldcraft-update crate (#231)
Versions ordered as SemVer, the GitHub Releases API reply read as hostile input, how this
copy is installed (macOS app or CLI, Windows folder, AppImage, tar.gz folder, .deb/.rpm) and
which release asset updates it, the checksum list's minisign signature checked against the
built-in release key, the asset's SHA-256 checked against the list, archives unpacked inside
the staging folder, and the swap planned and applied by renames with undo on failure. The
network and unpacking parts are native only; a mock release server drives the tests.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:48:30 -04:00
jyvikoandClaude Opus 5.5 4c18322a34 Engine: the half crank extrudes in blocks through mesh.hex.generate and mesh.cells
The extrusion's automatic cross-section blocks reach both commands; the
half crank arm, cut only on its symmetry plane, meshes all-structured
(sJ >= 0.37) with no seeds either way, and mesh.hex.generate's
description says so.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:26:30 -04:00
jyvikoandClaude Opus 5.5 02172f9a7e Mesh: swept cross-sections cut into blocks automatically
A region of an extrusion's (or revolution's) cross-section that no
structured scheme fits is now cut by straight lines into blocks that fit,
before interval assignment, so the cuts are curves of the arrangement and
their counts are assigned with the rest; nothing reaches the model
(hex::decompose, after Tam and Armstrong's medial-axis subdivision and Li,
Armstrong and McKeag's primitives).

Per region: several holes are separated by lines across the gaps between
neighbouring holes; a piece with one hole gets the O-grid layout whose
pieces score best (lines from the outer corners towards the hole or along
their bisectors, from the hole's corners, or along four directions from
the hole); pieces without holes are cut by a two-ply search over lines
from concave corners and fillets and from convex corners. Candidates are
scored by meshing the pieces with their schemes (smallest corner scaled
Jacobian). A region's blocks are kept when fair (at least 0.5) or no worse
than its free mesh, and cuts never end inside a seeded curve.

This replaces the cuts from reflex corners along an edge. Tests on an L,
a T, a plate with a hole, with a slot and with three holes, and the half
crank's cross-section: every region structured, the quads conforming and
the smallest corner scaled Jacobian bounded. The half crank arm now
extrudes in 12 structured regions with a hex20 minimum scaled Jacobian of
0.37 (the arm meeting the bosses 20 degrees off tangent bounds it) where
it had 2 free regions and 0.009.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:17:30 -04:00
jyvikoandClaude Opus 5.5 a0b4a69452 Mesh: the swept cross-section's quad mesh as its own module
The schemes, interval constraints, curve nodes and region meshes of a
multi-level sweep's cross-section move out of layered::mesh into
hex::section, with the curve placement (seeds, projection onto walls)
passed in, so the cross-section can be meshed on its own.
region::mesh_structured gives a scheme's mesh or None where it would fold,
without falling back to the free mesher. No change in behaviour.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:17:30 -04:00
jyvikoandClaude Opus 5.5 c83b803f63 web: the deployed browser build carries the PostHog key (#305)
Pages builds main with the repository secret POSTHOG_KEY, so web users are asked
the same consent question as desktop users; develop builds stay keyless.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:13:55 -04:00
jyvikoandClaude Opus 5.5 db80b75814 Docs: privacy.md lists every usage event; release builds get the key (#238)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:07:28 -04:00
jyvikoandClaude Opus 5.5 31bc886c7a Web: usage through fetch, preferences in local storage (#238)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:07:28 -04:00
jyvikoandClaude Opus 5.5 ed9ec7748d App: send usage batches to PostHog Cloud EU from a background thread (#238)
ureq over rustls (ring) with the OS certificate verifier. The project key
is the build-time FIELDCRAFT_POSTHOG_KEY; without it nothing is asked or
sent. FIELDCRAFT_TELEMETRY=0 turns it off; --control and CI runs only with
FIELDCRAFT_TELEMETRY=1. A failed post drops its batch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:07:28 -04:00
jyvikoandClaude Opus 5.5 79d583ea0f UI: ask once about anonymous usage; Help › Settings › Privacy (#238)
The first run asks whether to share anonymous usage (only when the host
says this run may), the answer and a random install id are kept in the
preferences, and Help › Settings… changes it. Commands are counted by
where they came from (ui, palette, script; mcp when the MCP server says
so), and cancelled forms and clicks on commands that cannot run yet are
recorded. The preferences go through a PrefsStore (a file, or the
browser's storage). Control channel: ui.settings, ui.prompt answers.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:07:28 -04:00
jyvikoandClaude Opus 5.5 162bb421f1 Engine: anonymous usage events, counted in Session::execute (#238)
fieldcraft_engine::usage holds the events docs/privacy.md lists, typed so
that only registry names (command ids, sample names, requirements) and
durations can go in them, a bounded queue that records nothing until the
host enables it, and the PostHog batch body. Session::execute counts each
top-level command of a session with an origin (ui, palette, script, mcp):
its id, kind, duration and error class, and the funnel's milestones and
opened samples. Nothing here touches the network.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:07:28 -04:00
jyvikoandClaude Opus 5.5 54f50f9518 Gallery: every example image captured again on the current build
All six docs/media/examples images at 1400 x 860 in the light theme, the
solved sample in the home view (the crank mirror from below, as its text
says), with the wireframe symbols, no SETUP card and the crank's planes
hidden; the Start card thumbnails cropped to the model from them.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:01:43 -04:00
jyvikoandClaude Opus 5.5 a0d11f1c22 Partitions hide their datum plane
geometry.partition hides the plane it cuts by (view state, saved with the
model): the cut shows as edges on the cells. The eye or selecting it in
the browser shows it again. The crank samples so open without their
planes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:01:43 -04:00
jyvikoandClaude Opus 5.5 dddb2d0584 Datums: drawn at the model's extent, never framed, the selected one shown
A datum plane is drawn as the model's extent within it (where it cuts the
model's box, in the plane's axes) plus a 5 % margin, a datum axis along the
model's extent plus the margin, instead of squares 1.2 diagonals wide.
Fit, Home, the standard views and headless pictures frame the model
without the datums (they stay inside the clipping radius). The datum
selected in the browser is drawn highlighted and on top, even when hidden.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 01:01:43 -04:00
jyvikoandClaude Opus 5.5 b19277ae7a Docs: crash reports, the log file, Report a Bug and privacy (#107)
docs/crash-reports.md: what is kept where, what a bug report carries
(and only when ticked), the CLI and commands for agents; README's
privacy note; the control protocol's new prompt answers.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:36:18 -04:00
jyvikoandClaude Opus 5.5 e06c85eb9e App: crash reports and a log file; Help › Report a Bug (#107)
The desktop app installs the crash reporter and a log file at start
(both in a temporary folder in --control and CI runs, #266), journals
the window's commands (background jobs included) and every control
request's method, and ends the journal on a clean exit. A start after a
crash asks: "FieldCraft closed unexpectedly", Report It… or Not Now.

Help › Report a Bug (also Tools › Help and ui.start
diagnostics.bug_report) shows exactly what the GitHub bug form will be
filled with: the version, the OS, the last error, and for a crash where
it happened. Share my session adds the replayable commands, Include the
model file the model. Open GitHub Issue opens the form in the browser
through the system's opener; files to attach are saved to
bug-reports/<time>/ and a long session goes on the clipboard. Nothing is
sent without that click and the user's submit on GitHub.

The control channel answers the new prompts (crashed: report, close;
report_bug: set, close) and ui.inspect shows the draft; it never opens
the browser.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:36:18 -04:00
jyvikoandClaude Opus 5.5 cb7e3a2711 CLI: fieldcraft-cli crashes [list|show|replay|dir|clear] (#107)
Lists, shows, replays and deletes the crash reports the desktop app
wrote. A replay opens the report's model, runs its commands with writes
confined to a scratch folder, and says whether the same internal error
happens again. Adapted from SolidCraft's solidcraft-cli crashes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:36:18 -04:00
jyvikoandClaude Opus 5.5 080ff048c6 Engine: crash reports, a log file and bug report drafts (#107)
crash: a panic hook and a locked per-run command journal (adapted from
SolidCraft's crash reports). A panic in a command becomes a report with
the model from before it and the command; one outside commands, or an
unclean exit found at the next start (abort, segfault, a crash in a
graphics driver, a kill: what #248 needs), a report with the model the
journal starts from (autosave, file or a copy) and the commands since,
the last control request and the dead run's log lines. Reports are
bounded, deduplicated by signature and replayable (crash::replay, with
writes confined to a scratch folder). The crash folder follows the
automation temp area in --control and CI runs.

logfile: a small size-rotated file logger behind the log facade.

bug_report: GitHub bug-form drafts. Version, OS and what happened by
default, home folder shown as ~; the command journal only with Share my
session, the model only with Include the model as well; a session too
long for the link becomes a file to attach and the clipboard text.

Commands: diagnostics.crash_reports, diagnostics.crash_report,
diagnostics.bug_report (Tools › Help, with its icon) and
diagnostics.crash_test (FIELDCRAFT_CRASH_TEST=1 only).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:36:18 -04:00
jyvikoandClaude Opus 5.5 3d51bde9dd io: the app's data folder, and confining writes alone
recovery::data_dir() names the per-user data folder (the temporary
fieldcraft-automation folder in automated runs) that the recovery, crash
and log folders live in. Roots::writes_within lets reads go anywhere
while writes stay in one folder, for replaying a crash report safely.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:36:18 -04:00
jyvikoandClaude Opus 5.5 d3ed8ffb05 Start card: every sample as a card with its thumbnail, flagship first, scrolling (#245)
The Start card and File › Samples show a grid of sample cards (thumbnail, name,
one-line description, the full text on hover) in a scroll area with a solid
scrollbar; Help › Samples keeps rows in the same order. The card stays clear of
the view cube and the navigation bar.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:06:20 -04:00
jyvikoandClaude Opus 5.5 072139353f Samples: 320 px thumbnails cropped from the gallery captures
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-15 00:06:20 -04:00
jyvikoandClaude Opus 5.5 08f669acad Browser and viewport: item menu, per-item eyes, Delete key with confirmation
Right-click on any browser item (body, surface, datum, partition, set,
material, section, step, boundary condition, load, reference point,
coupling, contact, fluid domain) opens its menu: Edit…, Hide/Show,
Isolate, Show All, Active in step (boundary conditions and loads) and
Delete…. A right click on a datum or a partition's cut face in the
viewport opens the same menu. Each item the viewport draws on its own
has an eye; hidden items are neither drawn nor picked. Datums and
partitions are selectable with properties. Delete… and the Delete key
ask first (the prompt lists what uses the item: a datum's partitions
are deleted with it, other definitions in use are refused); ui.prompt
answers it. Every entry runs the same command as the ribbon and forms.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:51:01 -04:00
jyvikoandClaude Opus 5.5 d7ec7f486a View state: hide, show and isolate viewport items, saved with the model
view.hide, view.show, view.isolate and view.show_all change which
bodies, surfaces, datums, boundary conditions, loads, reference points
and couplings the viewport draws. The hidden items are view state: the
manifest's `view` keeps them, a change marks the model unsaved, but it
is neither journalled nor undone and leaves results current. Items that
go away are forgotten. Headless pictures leave hidden items out; the
glyph builders take a filter. Tools > View > Show All claims the
commands.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:51:01 -04:00
jyvikoandClaude Opus 5.5 d9824de4a5 Engine: geometry.datum.point; planes and axes through datum points; datum edits and cascading delete
A datum point by coordinates, at a mesh vertex, a fraction along a
design edge, at a face's centre, or where a datum axis crosses a datum
plane. Datum planes take three datum points or a normal to a datum axis;
axes take two datum points. geometry.datum.edit, .axis.edit and
.point.edit redefine a datum of their kind (partitions cut again);
geometry.datum.delete with cascade also deletes the partitions that cut
by it. The ribbon shows Datum Plane, Datum Axis, Datum Point and
Partition, each with its own icon; edit and delete leave the ribbon
(still commands, claimed by their entries).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:51:01 -04:00
jyvikoandClaude Opus 5.5 fbaba5fecd Geometry: a face's area centroid and a point a fraction along an edge
Shape::face_centre and Edge::point_at place datum points on a design's
named faces and edges.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:51:01 -04:00
jyvikoandClaude Opus 5.5 0bf150633c Model: datum points, drawn as a small cross; each definition's users are public
Datum::Point joins planes and axes (checked like them, refused by
partitions with the kind named). The users each remove_* refuses with
(sets, materials, boundary conditions, loads, reference points, fluid
domains, datums) are public functions, so a caller can list them before
deleting.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:51:01 -04:00
jyvikoandClaude Opus 5.5 c1fd58c6ce release: version 0.5.8
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:33:11 -04:00
jyvikoandClaude Opus 5.5 90948833e0 Samples: the crank arm meshed structured from a partitioned half
crank-arm (and crank-mirror, kept as the same model under the name of
its mirror workflow) now cut the crank on its symmetry plane and the
half into seven cells: across the arm at x = 30, 60, 110 and 148, and on
two diagonals radial to the pocket's corner fillets. Every cell's
cross-section maps (four-sided blocks, the half O-grids round the bores,
triangle primitives at the pocket's corners); exact edge seeds give each
cut the same divisions on both sides, so mesh.cells merges all six cuts
and mesh.mirror reflects the half: 7 710 hex20, scaled Jacobian min 0.37
(was 0.17, and 0.009 for the swept mirrored half), edge ratio at most
7.0. Root σxx ±62.43 MPa against M/Z 62.6 MPa, the spindle carries the
1500 N. The test asserts the quality floor. Gallery images regenerated.

Fixes #247

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:17:20 -04:00
jyvikoandClaude Opus 5.5 73987bdbae Mesh: the triangle primitive for three-cornered sweep regions
A cross-section region with three corners (a fillet's quarter disk, a
pocket corner cut off by a partition) was left to the free all-quad
mesher. It now gets midpoint subdivision's triangle primitive (Tam and
Armstrong 1991; Li, Armstrong, McKeag 1997): lines from a centre to a
node on each side split it into three mapped blocks, the interior then
smoothed and every quadrilateral checked (a fold falls back to free).
Interval assignment gains dominance constraints (a side shorter than
the other two together) so the split always exists.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:17:20 -04:00
jyvikoandClaude Opus 5.5 7df37d4b15 README: the setup symbols screenshot redrawn with the wireframe symbols
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:14:40 -04:00
jyvikoandClaude Opus 5.5 52c12e2ad7 Setup symbols: wireframe glyphs per held degree of freedom, a constant size on screen
Boundary condition, load and interaction symbols are wireframes (thin lines,
no shaded cones), so they never hide the mesh or a contour plot:

- a crossed wire triangle per constrained translation, tip on the node,
  coming from outside the part where the set's faces say which side that is;
  a held rotation adds a smaller triangle behind it (the double mark); fixed
  shows all three axes, a symmetry plane only its normal;
- forces as open-headed arrows ending on the node, as long as four support
  units at one place and 2.5 over a set, so loads and supports stay in
  proportion (#78); moments as an axis line and a right-handed curved arrow
  (#79); contact and tie surfaces as small squares on their faces;
- a constant size on screen at any zoom (#80): the symbol unit is the size
  setting (View > Symbol size, `ui.set {symbol_size}`, 14 pt by default)
  times the model length a pixel spans, rounded to a quarter of a doubling;
- spread evenly on a grid a few symbol units across, so their density follows
  the screen, never more than 400 per definition;
- a layer of their own on the desktop (GPU and CPU): zooming rebuilds and
  uploads the symbols only, never the mesh's scene; framing leaves them out;
- colour per kind from the theme's tokens (a new contact colour), and the
  definition selected in the browser drawn in the selection colour, wider;
- depth-tested against the model with the line bias that keeps them on their
  own surface; a load at a reference point or a single node and a reference
  point's cross are drawn over the part (#78).

Headless pictures (MCP screenshot, `ui.render`) frame the camera first and
draw the symbols at the default size for it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:14:40 -04:00
jyvikoandClaude Opus 5.5 1fcc7acf84 render: lines drawn on top, any number of overlay layers, length per pixel
A scene's `on_top` lines are drawn after every depth-tested layer, over the
model, so a concentrated load at a point inside the part can be seen from any
side (#78). `render_layers` takes a list of overlays (the setup's symbols and
the selection highlight become separate layers), and draws every layer's
labels. `Camera::length_per_pixel` is the model length a pixel spans at the
target: the scale symbols of a fixed screen size are drawn at.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:14:40 -04:00
jyvikoandClaude Opus 5.5 3aab588b8b Navigation: orbit about the point under the cursor, zoom toward it, presets in the display menu
An orbit turns about the model point under the press (the pick ray's
first hit on the plot, the mesh or the design), or the centre of what is
drawn when it starts on empty space, keeps that point still on screen and
marks it while orbiting. The wheel and a pinch zoom toward the point
under the cursor; the wheel's delta is read as it arrives, so a notch
zooms in full in orthographic and perspective views alike (#273).

Mouse presets and reverse zoom come from SolidCraft (prefs.rs NavPreset,
NAV_PRESETS; MIT OR Apache-2.0): FieldCraft's own (left drag per the
navigation bar) stays the default, Ctrl + left drag from empty space
orbits in every preset. They are chosen in the display menu, settable
with ui.set and kept in prefs.json beside recent.json (the browser build
keeps them for the page's run); an older prefs.json still loads.

Fixes #262, fixes #273.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:02:34 -04:00
jyvikoandClaude Opus 5.5 f074e10442 Render: Camera::orbit_about turns the view about a pivot that keeps its place on screen
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 23:02:34 -04:00
jyvikoandClaude Opus 5.5 05abcec6c2 engine: the crank-arm conformity test checks all-hex20, not a platform-dependent count
The tet-free sweep gives 7274 hex20 on macOS and 7334 on Linux (#115), so the exact
count failed develop's CI after #272.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:49:31 -04:00
jyvikoandClaude Opus 5.5 012119e2c2 Remove the SETUP card: the browser tree shows the progress
The setup checklist card (#265), its show_setup view setting, the View
menu checkbox and the preferences file that only remembered it go; the
tree's marks and ghost rows replace them. Session::progress and
model.check's progress stay for agents. An old prefs.json is no longer
read.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:49:07 -04:00
jyvikoandClaude Opus 5.5 f5a84f5f74 Browser: setup state marks and ghost rows on the tree
Each top-level node that is a setup stage carries a mark at the right
of its row beside its count (done, next, blocked) with the reason and
place in its hover, and a node still lacking what its stage needs shows
a ghost row that starts the providing command. ui.inspect reports each
node's state as browser.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:49:07 -04:00
jyvikoandClaude Opus 5.5 910d437819 App: ask before a mesher replaces a body's mesh
The form's OK on a mesher whose body already has a mesh opens a Replace
mesh prompt ("CrankArm already has a mesh (7,274 hex20). Replace it?"),
naming sets that list the old labels and results that will be cleared.
Replace runs the form; Cancel leaves it open. The control channel's
ui.dialog submit is the OK button, ui.prompt answers replace or cancel.
Programmatic calls never ask. Fixes #267.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:45:24 -04:00
jyvikoandClaude Opus 5.5 0e4ea5e099 Meshers replace the mesh of the bodies they mesh, and only theirs
mesh.tet.generate and mesh.planar stacked a second mesh on a meshed body;
mesh.hex.*, mesh.hex.octree and mesh.cells wiped every body's mesh. All of
them, and mesh.mirror for its image bodies, now go through replace::clear:
the body's old elements and the nodes only they use go, other bodies stay.
The reply reports replaced {body, elements, nodes}, stale_sets (sets that
listed the old labels) and results_cleared; the command is one undo step.
replace::MESHERS names the bodies each mesher fills, and replacement()
says what a run would replace, for the app to ask first. mesh.block stays
additive: a body is built from several blocks. Refs #267.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:45:24 -04:00
jyvikoandClaude Opus 5.5 ed2541170a Model: remove one body's mesh, and list the sets that named its labels
Model::remove_body_mesh drops the elements that fill a body and the nodes no
other element uses, with their tags; Model::sets_using names the sets that
listed any of them by label. The base of per-body remeshing (#267).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:45:24 -04:00
jyvikoandClaude Opus 5.5 0bbdd463ea Cells: the blocks refusal names its problem as a schema Problem
develop did not build: #277's new refusal in mesh.cells passed text to
EngineError::param, which #276 changed to take a ProblemKind.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:45:24 -04:00
jyvikoandClaude Opus 5.5 65fb1112ab Engine: EngineError::Param carries the schema Problem at its field's path
The ~40 parameter checks give a ProblemKind instead of text; Session::execute looks the problem
up in the command's form (a group's missing field becomes that field's own path). Replies still
name keys; forms phrase the same problem by label. A test refuses every field of every form and
checks each refusal names a field of that form.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:42:19 -04:00
jyvikoandClaude Opus 5.5 de8b7b6321 Schema: one phrasing of a problem, by key or by label; Required by field kind
Problem::phrase serves both for_agents and for_form. A required name says "enter a name", a
path "enter a path", anything else "required" (no more "Autosave: enter an autosave"). New
kinds for what commands check (positive, range, list, name, path, invalid), and refine() to
look a command's problem up in its form.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:42:19 -04:00
jyvikoandClaude Opus 5.5 70c5c9b26f Ribbon: Hex Mesh and Partitioned Mesh replace the duplicate hex entries (#270)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:35:17 -04:00
jyvikoandClaude Opus 5.5 f4b6d1dbff mesh.cells: "default": "blocks" runs mesh.hex.blocks; labels Hex Mesh, Partitioned Mesh
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:35:17 -04:00
jyvikoandClaude Opus 5.5 2d5dfb2137 Forms: a one-line note per choice; the hex Method says what each needs
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:35:17 -04:00
jyvikoandClaude Opus 5.5 ad84a52ed2 Viewport: one framing rule for open, results, fit, home and standard views
Every fit frames what is drawn (the deformed shape and the setup's symbols
included) with Camera::fit_in in the part of the viewport the overlays
leave free: the results legend, the selection bar, the gizmo and the
navigation bar (framing::free_part). Standard views fit too, so a long
beam's front view keeps both ends (#147); fit and home leave the legend
clear (#76). The view frames itself in the home view when another model,
sample or design is opened (Session::opened; the same design opened again
after an edit keeps the camera) (#233), and when results come up with part
of the model out of sight (#223). ui.inspect's layout reports the legend.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:29:44 -04:00
jyvikoandClaude Opus 5.5 5758dc7911 Results: a static step first shows its contours undeformed
The first look at a solved static step draws the von Mises contours on the
true geometry; the deformed shape comes when asked for (Deformed,
results.deformed, results.animate), at the automatic scale. Frequency steps
still open on mode 1 deformed and explicit steps at true scale, as the
owner's notes in #77 ask; a 0.24 mm deflection drawn as 10 mm no longer
reads as a failed part.

Fixes #77

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:29:44 -04:00
jyvikoandClaude Opus 5.5 29af2aef91 Camera::fit_in frames a box's corners in part of the view
The distance and the target follow from the eight corners' exact
projections, for orthographic and perspective cameras and any direction of
view, so a long beam seen side on fills the width instead of a sphere
around it the height (#147). The free part is given as fractions of the
view, so overlays can be left out (#76); a margin is kept on each side.
Camera::fit is fit_in over the whole of a square view, and the headless
renderer fits to the picture's aspect.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:29:44 -04:00
jyvikoandClaude Opus 5.5 696b64c30f Session::opened identifies the open document
It changes when file.open, file.recover or file.sample replaces the session
and when geometry.open opens another design file or starts a new model; an
edit, a solve or reopening the same design keep it. The desktop app frames
the view on that change (#233).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:29:44 -04:00
jyvikoandClaude Opus 5.5 a0f48c7207 render: quadratic elements draw through their mid-side nodes; geometry edges over the mesh follow it
Zoomed in on the hex20 crank arm, the element edges and the shaded faces
did not meet (#260). The mesh is conforming; the display drew each hex20
face and edge as straight chords between corners while the mid-side
nodes lie on the curved surface (up to 0.44 mm off the chord at 3 mm),
and drew the geometry's face boundaries from the design tessellation,
off both.

Now a quadratic face is cut into triangles whose vertices are its
corners and mid-side nodes, each edge is drawn as its two halves, and
over the mesh the geometry's edges (and a hovered design edge) are drawn
as the element edges on geometry edges (`DisplayEdge::feature`), so
faces, edges, feature lines and the highlight share their vertices.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:13:15 -04:00
jyvikoandClaude Opus 5.5 9e38fe866d mesh: conformity check finds cracks (free faces, open edges, duplicate nodes, mid-side mismatches)
`quality::conformity` checks a solid mesh for cracks: the free faces must
close up (every corner edge on exactly two), no two distinct nodes may
coincide, and every element edge has one mid-side node. The hex tests
assert it for every method, hex20 included, and the crank arm sample's
7274 hex20 is conforming (#260: the disjoint lines are a display issue).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:13:15 -04:00
jyvikoandClaude Opus 5.5 c26b142630 Recover: each entry's age, and Discard all
file.recovery.list gives each entry's age; File › Recover shows name, age and counts, newest
first, with Discard all. docs/formats/fieldcraft.md states the folder and pruning rules.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:11:13 -04:00
jyvikoandClaude Opus 5.5 6bb3b04e60 App: a --control run keeps its autosaves out of the user's recovery folder
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:11:13 -04:00
jyvikoandClaude Opus 5.5 b1180bf987 IO: automated runs autosave into a temporary folder; stale entries prune themselves
Runs driven by automation (use_automation_dir) and CI use fieldcraft-automation/recovery in the
temporary folder unless FIELDCRAFT_RECOVERY_DIR is set. Taking a slot deletes left-behind
entries older than 14 days and those whose model was saved since.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 22:11:13 -04:00
jyvikoandClaude Opus 5.5 786121c21a UI: a setup checklist at the top of the browser
Once a model has geometry or a mesh, the browser shows the analysis stages from the engine's
checklist, each ticked, next, dashed (optional) or open; a click starts the stage's command and a
blocked stage's hover says what it waits for. The × hides it and View › Setup checklist shows it
again; the choice is kept in the per-user prefs.json.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:46:33 -04:00
jyvikoandClaude Opus 5.5 e223a6256a MCP: check_model describes the setup checklist it returns
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:46:33 -04:00
jyvikoandClaude Opus 5.5 fe557054ef Engine: the setup checklist (Session::progress), returned by model.check
The analysis stages in order (geometry, seeds & partitions, mesh, material, section, sets,
boundary conditions, loads, step, solve, results), each done, next, todo, blocked or optional,
with the command that does it. Built on the command requirements; Solve also reads the model
check's first error.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:46:33 -04:00
jyvikoandClaude Opus 5.5 fc85550534 UI: a form's label column fits its longest label, long labels wrap, fields fill the row
The column is measured from the form's labels, at most 45 % of the form's
width; a longer label wraps onto two lines (the column widens only when
two are not enough) and its required and help marks follow it inside the
column. Fixes labels cut at a fixed 112 pt ("Hardening modu…").

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:45:06 -04:00
jyvikoandClaude Opus 5.5 3d8864c313 UI: browser rows keep their count and bias visible, long names elide in the middle
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:45:06 -04:00
jyvikoandClaude Opus 5.5 2443746b35 UI: status messages elide in the middle, the command line's placeholder fits its width
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:45:06 -04:00
jyvikoandClaude Opus 5.5 8b12cdd5c7 UI: field help on hover and an (i) mark, units without a unit system, X/Y/Z captions
A form row's label and its drawn (i) mark show the field's help and unit
on hover. With no unit system a number shows its dimension ("[stress]")
and the form says where to set the system (Properties › Units). Vector
components are captioned (X, Y, Z, or the field's own names such as E1).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:29:07 -04:00
jyvikoandClaude Opus 5.5 6b9e669c7e Engine: every form field says what it means, and shows its unit or dimension
Field gains `help`: one or two sentences of our own wording with a
typical value (yield stress: the initial yield, 276 MPa for 6061-T6,
250 MPa for mild steel; hardening modulus explained; and so on), set
next to each field in cmd/forms.rs. Dim::name gives the dimension
("stress"), Field::unit_label the unit or, with no unit system, the
dimension in brackets, and Field::help_text the tooltip with the unit or
where to set the unit system. The JSON Schema description carries the
help too. A test checks every form field has help and every dimensioned
field shows a unit or dimension in every system.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:29:07 -04:00
jyvikoandClaude Opus 5.5 eb66b86e26 Engine: the meshing commands pass the design's edge seeds to the meshers
mesh.tet.generate, mesh.hex.*, mesh.hex.blocks and mesh.cells give every
seed on the open design to their meshers, which match them to their own
curves by position, so a seed on an edge two cells share holds in both. The
tet reply gives, per seed, the mesh edges along it. Docs and the roadmap
say what is honoured where.

Fixes #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:23:05 -04:00
jyvikoandClaude Opus 5.5 f78680e091 Mesh: the tet mesher divides a seeded edge at the seed's nodes
The seed's spacing becomes size-field samples the field grades from, and
after coarsening its nodes are put on the feature curve and pinned, so the
edge has exactly the seeded divisions at the seeded spacing. The quality
report gives, per seed, the boundary edges along it.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:23:05 -04:00
jyvikoandClaude Opus 5.5 13b506150c Mesh: the hex meshers honour edge seeds
A seed is matched to a mesher's curve by position (ends and middle within the
two tessellations' tolerance), which also orients its bias. Interval
assignment holds exact seeds, raises minimums and takes free seeds as goals;
an equality, loop minimum or parity it cannot meet with the exact seeds held
names them. Seeded curves get the seed's spacing. Extrude and revolve seed
cross-section curves and the layers of a level interval; sweep and mapped
seed their B-rep edges; cells seed the shared edge once, so both cells keep
it and stay conforming.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:23:05 -04:00
jyvikoandClaude Opus 5.5 bdd84a491d Schema: a problem carries its field and a structured kind, phrased by label or by key
schema::Problem now holds the &'static Field it is about and a ProblemKind
(Required, NotANumber, Below, Above, NotAChoice, UnknownVariant, ...).
Problem::for_form phrases it by the field's label, Problem::for_agents by
its JSON path, from the same data. The form no longer rewrites problem
text by string matching or re-walks the field tree; field_at moves to
schema.rs for the engine's own Param errors. The forms test matches
EngineError::Param { key, .. } directly.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 21:09:22 -04:00
jyvikoandClaude Opus 5.5 904dc3381b UI: form problems name fields by their labels, and OK waits for required fields
Every problem reads with the field's label ("Material: choose a material")
beside its field, once it was edited or OK was tried; a click on the greyed
OK shows them all, and its tooltip lists them. A parameter the command
refuses shows beside its field, by its label, in the form and status bar.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:58:45 -04:00
jyvikoandClaude Opus 5.5 9dcbb9c9a6 Engine: a parameter error carries its key and problem (EngineError::Param)
A form can show the problem beside the field it is about. The message
still names the key, as before, for agents.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:58:45 -04:00
jyvikoandClaude Opus 5.5 8df4ea040c UI: seeds as ticks along their edges, Mesh › Seeds, and seeding the selection
A seed draws a tick at each node it puts on its edge, so a bias shows as ticks
crowding toward an end. The browser lists the seeds under Mesh; a click
selects the seeded edge. The Seed Edges form takes the selected edges with
Use selection. The toolbar entry is ready (its icon renamed with it).

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:58:17 -04:00
jyvikoandClaude Opus 5.5 e8cada94d0 Engine: mesh.seed.edge, mesh.seed.clear and mesh.seed.list
Seed design edges by their persistent names with a count or a size, a bias
and a constraint; one undo step each, saved with the model. Seeds on edges a
reopened design lacks are reported as stale.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:58:17 -04:00
jyvikoandClaude Opus 5.5 f8e0be970e Model: edge seeds — a count or size per edge, bias and constraint
A seed names a design edge and asks for a number of divisions or a target
length, optionally graded geometrically toward the start, the end or both
ends, kept exactly, as a minimum or as a free goal. Seeds are saved with the
model; Seed::spacing gives the node positions meshers and the viewport share.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:58:17 -04:00
jyvikoandClaude Opus 5.5 94c98e0c0b UI: pick a design's edges by their persistent names
The viewport keeps each drawn body's B-rep edges; with Select: Edges a click
picks the nearest visible one (before and after meshing), hover and the
selection draw it wide, and multi-select works as for faces.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:46:05 -04:00
jyvikoandClaude Opus 5.5 ffb71de6e1 Engine: geometry.bodies lists each body's named edges
Each edge with its persistent name and length. A reference to an edge name
selects its numbered pieces (`~k`), as a face name selects its `#k` pieces.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:46:05 -04:00
jyvikoandClaude Opus 5.5 a11b971dae Geometry: persistent edge names, after the faces each edge separates
A design edge is named by its faces' persistent names, sorted and joined by
`|`, numbered `~k` by position when several separate the same faces. The
names follow the faces through a design edit and a partition replay.

Part of #230.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:46:05 -04:00
jyvikoandClaude Opus 5.5 6bfa0b410a UI: a reference field is always a drop-down; with nothing to pick it offers Create …, and one candidate starts picked (#215)
A reference field (a material, set, step…) with no items was a free-text box; it is now a greyed
drop-down ("No materials yet") with a button that opens the command making one (the
collection's requirement provider). That form comes back to this one on OK, with the new item
picked, or as it was on Cancel. A required reference with one candidate, or with the browser's
selection of its kind, starts picked when the form opens, so the fields and the JSON view agree.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:34:24 -04:00
jyvikoandClaude Opus 5.5 29e7a4ca31 Engine: a reference collection names the requirement (and its provider) that makes an item
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:34:24 -04:00
jyvikoandClaude Opus 5.5 794122a302 MCP: the partition, cell and mirror commands in the instructions and docs
The mesh tool runs mesh.cells (method cells) and mesh.mirror (method mirror).
The server instructions and docs/mcp.md list the partition, per-cell, mirror
and symmetry commands and point to the crank-mirror sample.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:23:04 -04:00
jyvikoandClaude Opus 5.5 860707efa6 Sample: crank-mirror, the crank arm meshed from a half and mirrored
file.sample crank-mirror partitions the crank arm on its symmetry plane, sweeps
one half all-hex (hex20) and mirrors it into one conforming mesh, then defines
the crank-arm model. Samples may now run preparing commands (datums,
partitions) before their script sees the bodies. Its test checks the halves
alike and joined, every element valid and the sets on both halves; the heavy
test solves it: sigma_xx at the root -62.43 / +62.43 MPa against M/Z = 62.5.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:23:04 -04:00
jyvikoandClaude Opus 5.5 a84a8339ae UI: command replies read as one line, the whole reply in the context panel
Commands run from the UI whose reply had no `summary` printed their raw
JSON into the status bar (Check Model, Step Table, Model Info, Bodies,
Find Contacts, the solve steps...). The status bar now gets one sentence
for every reply, and a reply with more to it opens in the context panel:
Check Model as its diagnostics with severity marks and a button to each
fix, the Step Table as a grid of loads and BCs against steps, any other
reply as labelled rows and tables. Check Model and the Step Table follow
the model as it changes. Programmatic replies stay JSON.

Fixes #217

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:18:04 -04:00
jyvikoandClaude Opus 5.5 f89545a66a Engine: model.boundary_condition.symmetry on a datum plane
A symmetry condition (zero normal displacement, zero rotation about the
in-plane axes) on a datum plane normal to X, Y or Z, with its set given or made
of the open design's faces on the plane (a partition's cut faces), so half a
symmetric part is solved as a half. Half a 40 mm bar cut at its mid-width, with
the condition and half the load, bends as the whole bar to 2e-12 with half the
reaction. Inclined planes are refused as not supported yet.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:08:42 -04:00
jyvikoandClaude Opus 5.5 0f99a1c687 Engine: mesh.mirror reflects meshed bodies across a plane
The bodies' elements are reflected across a datum plane (or a point and
normal) with fieldcraft_mesh::mirror and joined on it. The image is tagged as
the design's body at the reflected place (found by volume and centroid, or
named in `as`), its faces with that body's faces under them, so sets by face
name reach both halves. Half a bar meshed hex8 or hex20 and mirrored has the
whole bar's node and element counts and volume, and its bending deflection to
1e-12; tet4 and tet10 mirror valid and conforming.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:00:59 -04:00
jyvikoandClaude Opus 5.5 9fc8e27933 Geometry: the nearest face of a body to points
Shape::nearest_faces finds, for each point, the face whose tessellation lies
nearest: how a mirrored mesh finds the faces of the body it is the image of.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:00:59 -04:00
jyvikoandClaude Opus 5.5 b73a3bb976 Mesh: mirror elements across a plane, right-handed, sharing the plane's nodes
fieldcraft_mesh::mirror reflects a region's tet4, tet10, wedge6, hex8 and hex20
elements across a plane. The copies list their nodes in reflected order (corners
swapped, mid-side nodes following their edges), so every copy is right-handed;
reflected_faces and reflected_edges give the copy's face and edge over the same
nodes. Copies of the nodes on the plane are merged with the originals through
join::merge_coincident, so the region and its image form one conforming mesh.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 20:00:59 -04:00
jyvikoandClaude Opus 5.5 7d689fb6a2 UI: the CPU viewport image never exceeds the texture size limit
A Retina window 2300 pixels wide drew its CPU viewport (the offscreen capture's
path) into a 2300 x 1412 texture, past egui's 2048 limit, and the app aborted.
The image is now scaled to fit the context's max_texture_side and drawn
stretched, as an image over the renderer's own limit already was. Fixes #237.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 19:50:37 -04:00
jyvikoandClaude Opus 5.5 0b6b6c966c Engine: mesh.cells meshes a partitioned part cell by cell
Each cell (a body of the open design) gets its own method: structured hex where
one fits (auto falls back to tet and says why), or tet. All go into one mesh.
Cells on either side of a partition are joined: where their element faces
match node for node the nodes are merged; otherwise (hex next to tet, or
different divisions) the two partition faces get a set each and a dual mortar
tie, the reported hex-tet transition. A 40 mm bar cut in half, one half hex,
the other tet, stretches F L / (E A) to 1e-6.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 19:50:37 -04:00
jyvikoandClaude Opus 5.5 1fbb5581ea Mesh: merge coincident nodes, keeping tags and sets on the survivor
Model::merge_nodes replaces merged nodes by their survivors in elements (node
order kept, so face and edge tags hold) and in set node lists, moves a vertex
tag to an untagged survivor, and refuses a merge that would collapse an
element. fieldcraft_mesh::join::coincident_nodes / merge_coincident find the
nodes within a distance of a lower-labelled survivor, deterministically.
Meshing cell by cell and mirroring use them; node merging commands will too.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UhhZMBTviDQxZvwnyP89fo
2026-10-14 19:50:37 -04:00