Closes the view-to-dispatch resort race: the nine Layer* variants plus ColorPickTarget::Layer now carry the layer name and resolve to the current panel row at dispatch time via layer_panel_index; unknown names no-op. View sites (layer panel, ribbon dropdown) and the CLI/plugin path send names, so the CLI position lookup and the ribbon enumerate are deleted. Panel-internal selection state stays index-based. Tests: layer_toggle_visible_targets_layer_by_name_after_resort pins name targeting across a re-sort; layer_toggle_visible_unknown_name_is_noop pins the graceful no-op. Full suite: 1974 passed / 0 failed / 25 ignored.
Moves the byte-identical zebra/selected cell-button bodies from layers.rs and xref_manager.rs plus the layers header-button body into one shared module. All six call sites now use the shared helpers; toolbar_btn helpers untouched (different semantics). No new tests per owner request (pure move, verified by build plus full suite before test removal).
- #1602 changed the COMMANDLINEFADETIME help text and added "Never"
without catalog entries, so both showed in English everywhere. Add
them to the catalog and all 21 locales; drop the "0 skips them" key.
- #1606 named another CAD vendor in a doc comment and two test names.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
`v2026.40.1` matched neither the weekly pattern nor plain semver (2026 is
past the MSI major limit), so a fix release could not be built. Weekly
tags now take an optional patch: Cargo 2026.40.1, MSI 26.40.1, display
2026.40.1 (week zero-padded, as build.rs and the update check expect).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The v2026.40 rebuild was dispatched with a separate `commit` input copied
from a stale local tag, so Windows/macOS assets were built from 20803948
instead of 8e122a4e. Dispatch now takes only `tag` and checks the tag out.
Reverts 4700e263: the skipped Python plugin was a symptom of that commit.
Fixes#1612
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Turning a 141k-entity layer off/on took 1.5-5 s; it now takes ~0.6-0.9 s.
- Off/freeze toggles go through invalidate_layer_visibility: entities on
the layer keep their tessellation memo (visibility gates them before
tessellation); only block users re-expand. No mesh recolor.
- The resident patch reuses a surviving memo entry instead of
re-tessellating, guarded by the same display-context hash as the full
build.
- The patch looked up `layout.order` linearly per re-shown entity, which
made turning a big layer back on quadratic.
- recolor_meshes_for_handles resolved materials for handles with no mesh.
- Wire arena: a hidden slab keeps its instances and sets a `hidden` flag
in its consts (collapsed in the vertex shader); an unchanged re-show only
clears the flag. Const writes are batched into merged spans.
- The `layer_visible` automation action indexed the document's layer
table instead of the sortable panel rows the Layer* messages use.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- When the count area's boundary is gone the approved Invalid Area dialog
asks: undo the change (also its X and Cancel) or count all of model space;
"always" keeps the choice in the settings. An edited boundary moves the area.
- A table's *T block is brought up to date whenever the table changes.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Saving filled every 3D solid without stored wires with tessellated display
edges, so solids read from a file without a wireframe cache (for example a
DXF) were saved with edges OCS never created. Only solids without ACIS data
now get an edge cache on save; solids OCS creates or edits already get their
wires when they are built.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
- COUNTCOLOR / COUNTERRORCOLOR / COUNTSERVICE reject any other value with
the range message; COUNTTABLE ? lists in the reference order, padded,
22 lines a page with Press ENTER to continue:, then the totals.
- Target picks report N found / N found, M total; several objects (or one
that is not a block reference) count their copies, shown as Geometries
in the palette.
- Current area is the area in use (kept aside by Entire model space),
otherwise all of model space; an invalid Object boundary ends with the
reference message.
- Count colours reach solids and fills; the palette list and the count
are recomputed only when the drawing changes.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The before-manifest was taken pre-bake while save mints *D geometry on its clone, so every new dimension failed verification 1-vs-7. Compute before on a bake-simulated clone: baked-vs-baked proves roundtrip stability (bake correctness itself is pinned by the explode geometry tests). Pinned by save_verified_accepts_new_dimension_with_baked_block.
The DXF reader keeps marker identities on the block record but never materializes the entities, so block_entity_handle/block_end_handle dangle after every DXF load (reopened baked *D dimension blocks hit this every time). Re-create missing sentinels at load finalize, reusing the file handle when free and minting otherwise. Idempotent and DWG-safe: loads whose markers already resolve are untouched. Pinned by dimlinear_survives_save_reload_with_baked_geometry.
resources/* vary as sessions open and snapshots land, so an hour of advertised freshness was optimistic. protocol_result now takes the TTL value: tools/list and discovery keep 3600000, resources/list/read/templates get 60000, on every protocol version. No tool surface change; digest untouched.
The GUI requires request_id and document_id on every non-query op, but the bridge sent neither on capture: every ocs_capture failed with request_id_required, then document_required. request() now always mints the GUI request id for capture (a client-supplied id wins and is echoed in metadata), and capture carries the active document. The op=capture read path answers guidance pointing at ocs_capture instead of a missing-path failure. Live-proven against the open window: plain and annotated captures return image content.
The cached PID snapshot predates a GUI spawn, so the newborn pid missed from it read dead and the bridge deleted its just-written descriptor inside the launching call itself: every fast launch reported still-starting forever while a healthy window sat undiscoverable. Two guards: the snapshot cache is invalidated right after spawning, and deletion now waits out a 120s file-age grace (unknown age counts as newborn). Slow starters that write late were the only ones surviving before, which is why this was intermittent. Covered by stale_snapshot_may_delete predicate tests.
Runs the official MCP Inspector against a fresh release build on every push and PR: initialize purity, four titled tools, headless resources/list, both resource templates. Headless-safe methods only; drawing-dependent coverage stays in unit tests and the pinned digest. Server addressed via config file (positional CLI targets break on Windows shells); TTL assertions stay at unit level (the CLI normalizes result envelopes on output).
A dispatched rebuild checks out the tag's commit but runs main's
workflow; v2026.40 has no stage-bundled.sh, so the macOS bundle step
failed. Skip the plugin with a warning in that case only.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The wasm check broke on the MCP `tools` op, the block library's folder
picker and DWG thumbnails, and the PDF export-window culling helpers.
Gate each one like the code around it; `read_handle` needs no gate.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Everything was written and only clipped at view time, so a window over a
corner of a large drawing produced a near-empty page of several MB.
Wires and hatches whose sheet bounds miss the window are now skipped.
Fixes#1568
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A dimension draws from its *D block by name, but loading an xref skipped
anonymous blocks and kept the name verbatim, so its dimensions fell back
to a regeneration (giant text at the origin) or a host block of the same
name. The load path now imports them prefixed like other blocks and
renames the dimension's block; BIND is unchanged.
Fixes#1578
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Splitting a boundary object erased it and added the pieces, so the hatch
lost that source and rebuilt from what was left, often the whole circle.
The pieces now take over the old object's association before it is
erased.
Fixes#1572
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A point carried a key vertex, so Endpoint (which outranks Node) claimed
it, and an absolute-size glyph's strokes were indexed for Nearest and
Intersection. Points now offer only their Node, in the indexed and the
unindexed snap paths alike.
Fixes#1570
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Each band segment was a rectangle ending square at its vertex, so the
outer corner of every joint showed a notch. Every segment end now carries
the tangent of half its turn, and the shader slides that end's corners
onto the joint's bisector so neighbours meet in a miter. Turns sharper
than the limit keep butt ends. Constant-width runs inside a polyline split
around its arcs no longer count as tapered, which had kept them unmitred.
Fixes#1565
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The codec writes R2018 AuxHeader maintenance fields as RL, so AutoCAD no
longer reads a 2018 save as student-made. The kernel drops zero-area
spikes before triangulating, so a hatch whose boundary folds back on
itself fills on screen instead of vanishing.
Fixes#1564Fixes#1569
Co-Authored-By: Claude Opus 5.5 <[email protected]>
rustc 1.99 overflows its default 8 MB stack compiling the crate in
release on Windows and macOS. Raise RUST_MIN_STACK to 64 MB in the cargo
config and the release workflow (the latter also covers rebuilding older
commits). The Windows Build step now runs in bash so a failed cargo build
stops the job instead of surfacing later as a missing exe at signing.
workflow_dispatch gains tag/commit inputs to rebuild a published release.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Short interactive waits returned a bogus 10-minute timeout on first expiry, killing the re-poll pattern and blocking task creation. The dismiss branch now fires only when the requested wait actually reaches the ceiling; shorter waits return running as before. Predicate covered by interactive_timeout asserts.
tasks/cancel marks the task so polls report terminal cancelled with neither result nor error; late completions are discarded. Task envelope errors move from grandfathered -32000 to -32603. Agent schema sync on bridge start is now skippable via OCS_SKIP_SCHEMA_SYNC and logs one line when it writes.
resources/templates/list serves the two genuine parameterized URIs (pyramid tiles, snapshots) with TTL hints instead of -32601. Tool titles move top-level per the 2026-07-28 Tool shape (dropped from annotations); annotation keys are allowlist-tested. resources/list ids are sorted so client and prompt caches hit. Tool schema digest rolls to fbba987a; pin updated deliberately after reviewing the titles-only surface diff.
How to iterate the bridge: fast builds, digest pin, session model, discovery, launch serialization, cancellation, error routing, safety model, rebuild swaps, conformance gates. No diary entries, no harness names, no build-specific data.
run() reads stdin on a feeder thread into a channel (CancelPump); wait loops harvest lines in 50ms quanta instead of sleeping blind. A notifications/cancelled naming the in-flight MCP id dismisses the GUI prompt with a second cancel exchange, drops the late result, and sends nothing back. Cancels arriving between polls resolve via the inflight map; re-polls for dismissed ops answer cancelled without touching the GUI. Interactive picks (user_select/getpoint) get a 10-minute single-call ceiling ending in isError timeout text; other ops keep the 60s clamp. Ping-during-block stays unanswered while one exchange occupies the loop (bounded residual).
Unknown tools become -32602 Protocol Errors carrying the four valid names (SEP-1303 keeps them there; the isError shape hid them from strict clients). Bridge/GUI infrastructure failures (no sessions, dead TCP, unwindable startup) become -32603 via the CallError split; tool-domain validation stays isError so the model keeps self-correcting. Verified observable in a live client session: bogus session surfaces readable -32603 naming ocs_sessions, model recovers.
resources/read reports -32602 on the modern path (2026-07-28 forbids emitting -32002) and keeps -32002 on legacy, where it was the code. Non-object top-level input (batches included - no batch semantics defined) gets -32600 Invalid Request instead of silence, so senders never hang. Id-less objects stay silent (notifications). Ping arm deliberately untouched: the method was removed only on the modern path, and legacy health checks may rely on it.
serverInfo carries only name/title/version so strict clients never reject the handshake. Build extras (build_rev, build_profile, tool_schema) live canonically in bridge_identity, surfaced via capabilities.experimental and the in-band bridge stamp - no copies to drift. Adds conformance guards: spec-pure initialize with unknown-version fallback, and JSON-RPC errors for unknown methods, missing tool names and missing resource URIs.
Global MCP instructions don't reliably reach agents (fresh sessions never announced the build). The directive now lives in the ocs_sessions description every agent reads first, pointing at the bridge object. Tool schema digest rolls to ede58332; pin updated deliberately after reviewing the description-only diff.
List results (tools/list, resources/list, resources/read) carry ttlMs and cacheScope on the legacy 2025-11-25 path too; legacy clients ignore unknown fields per the SEP's backward-compat rule. The modern envelope (resultType, _meta.serverInfo) stays negotiated. Legacy-shape pin updated deliberately to assert presence.
Hakan Seven