feat(wipe-tower): BBS no-sparse clearance and safe descend (Orca #15485)

[aceRage]

Description

Port of OrcaSlicer #15485 (MERGED, 7c742e09c7) onto Edge main, plus a labeled second commit of the remaining Orca #15441 Z-restore.

Makes No sparse layers usable on Bambu-style prime towers:

• Toolhead descends to the compacted tower only when already over it (append_tcr defer-descend + in-tower NC travel).
• Pre-slice + post-tower clearance validation (compacted_wipe_tower_clearance_*) with plate highlight polygons.
• Auto-disable / message when smooth timelapse needs an every-layer tower.
• New extruder_clearance_dist_to_rod (default-off / unused until a profile sets it).
• Catch2 wipe-tower compaction/clearance tests; GLCanvas live by-layer preview (counterpart of sequential by-object clearance).
• fix: fix no sparse layers during a toolchange OrcaSlicer/OrcaSlicer#15441: change_filament_gcode runs at real layer Z, then Z returns to the tower.

Fixes Orca OrcaSlicer#15381 / Orca OrcaSlicer#11703 class of collisions. Option stays default-off; default profiles should match baseline G-code.

Commits

1. feat(wipe-tower): BBS no-sparse clearance and safe descend (Orca #15485) — Print clearance, GCode descend/travel, WipeTower+WipeTower2 helpers, GLCanvas/PartPlate preview, config/UI/timelapse, Catch2.
2. fix(wipe-tower): restore layer Z around change_filament_gcode (Orca #15441) — placeholders + restore/deretraction around toolchange. Verified against Orca fix: fix no sparse layers during a toolchange OrcaSlicer/OrcaSlicer#15441 head d1b31fee (OPEN, mergeable_state: dirty; the final-purge commits were already reverted upstream, so this is the remaining Z-restore only).

OrcaSlicer#15441 coordination

Orca OrcaSlicer#15441 is still OPEN and not green (CI is effectively snyk-only). Policy was: re-diff every GCode/WipeTower hunk both PRs touch; include Z-restore as a second commit if it composes safely; otherwise land OrcaSlicer#15485 alone and document residual multi-toolhead risk.

Included. The remaining OrcaSlicer#15441 patch is small (+58/−1 upstream, 3 files) and composes with OrcaSlicer#15485:

Hotspot	OrcaSlicer#15485	OrcaSlicer#15441	This PR
GCode.cpp append_tcr initial Z	defer_compacted_descend — do not drop onto the model	will_go_down gates restore/deretraction around change_filament_gcode	Combined: will_go_down && !defer_compacted_descend for the immediate drop; restore is a no-op when still at layer Z; deretraction is the in-tower descend after custom G-code
GCode.cpp append_tcr after toolchange	compacted_below_object descend injected into [filament_start_gcode]	[deretraction_from_wipe_tower_generator] after [change_filament_gcode]	Complementary: OrcaSlicer#15441 is the toolchange-path descend; OrcaSlicer#15485 start-filament descend is a backup if the writer is still above the tower
GCode.cpp append_tcr2	Unchanged Type2 descend (Orca OrcaSlicer#15485 also left this path alone)	Restore layer Z before set_extruder when will_go_down && enable_prime_tower	Ported; Type2 already had will_go_down + [deretraction_from_wipe_tower_generator]
WipeTower.cpp toolchange_Change	No placeholder change	[restore_layer_z_before_toolchange] / [deretraction_from_wipe_tower_generator] around [change_filament_gcode]	Ported
WipeTower2.cpp toolchange_Change	No placeholder change	[restore_layer_z_before_toolchange] before [change_filament_gcode]	Ported; existing deretraction placeholder kept

If Orca OrcaSlicer#15441 later grows past d1b31fee (e.g. the reverted final-purge work returns), re-diff before merging this.

Edge adaptations (not a blind cherry-pick)

• No wrapping_detection on Edge — omitted from the sparse-skip rule and from Catch2.
• Estimated footprint uses wipe_tower_mesh_data.bottom or width/depth (no wtd.bbx / rib_offset). Extrusion-to-bed is rotate+translate only.
• Sequential by-object clearance gizmo (Edge fix(print): sequential clearance Kahn sort + above-rod Y extent (#630, #793) #4 / Snapmaker Fix sequential print collision check depending on object list order Snapmaker/OrcaSlicer#630/Sequential print: only material above height-to-rod can collide with the beam (stacked on #630) Snapmaker/OrcaSlicer#793) is unchanged: sequential_print_clearance_valid, update_sequential_clearance, can_sequential_clearance_show_in_gizmo are not renamed. Compacted-tower preview is a parallel hook that reuses the same draw path (set_sequential_print_clearance_polygons).
• Both WipeTower (BBL append_tcr) and WipeTower2 (append_tcr2) paths are updated.
• H2C rack still injects nozzle-change into [filament_end_gcode] before the new restore placeholder.
• MixedFilament flush gates, MQTT, paint-depth, ImageMap, and CLI were not touched.

Screenshots/Recordings/Graphs

N/A in this environment (no GUI session). Expected plater behaviour: by-layer compacted-tower keep-out rings beside the existing sequential by-object gizmo; unsafe layouts get the shared clearance message and slicing is blocked.

Tests

• Catch2 additions in tests/libslic3r/test_wipe_tower.cpp ([WipeTower][NoSparseLayers]): sparse-skip rule, compacted Z, keep-out zone, per-object rise limits, ring/outline agreement, brim padding.
• Catch2 status: not executed here — this environment has no CMake/deps build tree (CMakeCache.txt absent). Please run cd build && ctest --output-on-failure -R libslic3r_tests (or ./tests/libslic3r/libslic3r_tests "[NoSparseLayers]") after a normal slicer build.
• Upstream fix: fix no sparse layers during a toolchange OrcaSlicer/OrcaSlicer#15441 has no extra Catch2; Z-restore is G-code ordering and needs a multi-filament no-sparse slice to confirm change_filament_gcode at layer Z.

Accept criteria

• Default profiles (no-sparse off): G-code / behaviour match baseline.
• No-sparse on + safe layout: tower skips sparse layers; descend only after travel over tower.
• Unsafe layout: pre-slice error + plate highlight; slicing blocked with shared message.
• Smooth timelapse: setting forced off / explained.
• Sequential by-object clearance gizmo still works.
• Catch2 wipe-tower tests green (see status above).
• fix: fix no sparse layers during a toolchange OrcaSlicer/OrcaSlicer#15441 included: toolchange custom G-code is restored to real layer Z then returned to tower Z (WipeTower + WipeTower2). Still needs a real multi-toolhead no-sparse slice to confirm; do not treat this as printer-validated.

Residual if this second commit were dropped: Type2 / !is_finish_first can still run custom G-code at tower Z and sweep through taller parts. With the second commit landed, that is the remaining risk only if a later OrcaSlicer#15441 revision diverges from d1b31fee.

  

[github-actions]

❌ Documentation validation failed

🔗 Link Validation Errors

📄 doc/developer-reference\Built-in-placeholders-variables.md:

• Line 93: home#process-settings - Fragment does not exist
• Line 93: home#material-settings - Fragment does not exist

📄 doc/developer-reference\How-to-wiki.md:

• Line 44: home#material-settings - Fragment does not exist
• Line 45: home#process-settings - Fragment does not exist
• Line 46: home#prepare - Fragment does not exist
• Line 47: home#calibrations - Fragment does not exist
• Line 50: home#process-settings - Fragment does not exist
• Line 50: home#support-settings - Fragment does not exist
• Line 50: home#others-settings - Fragment does not exist

📄 doc/Home.md:

• Line 14: Precise-wall.md - File does not exist
• Line 17: extrusion-rate-smoothing.md - File does not exist
• Line 33: adaptive-pressure-advance.md - File does not exist

📄 doc/Tab.cpp:

• Line 2376: quality_settings_z_contouring - File does not exist
• Line 2377: quality_settings_z_contouring - File does not exist
• Line 2378: quality_settings_z_contouring - File does not exist
• Line 2379: quality_settings_z_contouring - File does not exist
• Line 2380: quality_settings_z_contouring - File does not exist
• Line 2443: fill-patterns - File does not exist
• Line 2444: fill-patterns - File does not exist
• Line 2447: fill-patterns - File does not exist
• Line 2677: multimaterial_settings_prime_tower#brim-chamfer - Fragment does not exist
• Line 2678: multimaterial_settings_prime_tower#brim-chamfer-max-width - Fragment does not exist
• Line 2690: multimaterial_settings_prime_tower#wall-gap - Fragment does not exist
• Line 2707: multimaterial_settings_ooze_prevention#delta-temperature - Fragment does not exist
• Line 2719: multimaterial_settings_advanced#paint-depth-mode - Fragment does not exist
• Line 2720: multimaterial_settings_advanced#paint-depth-walls - Fragment does not exist
• Line 2721: multimaterial_settings_advanced#paint-depth-distance - Fragment does not exist
• Line 2725: multimaterial_settings_advanced#paint-sparse-infill - Fragment does not exist
• Line 2729: multimaterial_settings_advanced#paint-depth-solid-interfaces - Fragment does not exist
• Line 2784: others_settings_fuzzy_skin#skip-fuzzy-skin-on-overhangs - Fragment does not exist

🖼️ Image Validation Errors

📄 doc/developer-reference\How-to-create-profiles.md:

• Line 142: [Markdown] Alt text "Help menu" ≠ filename "go-to-configuration-folder"
• Line 144: [Markdown] Alt text "Delete system folder" ≠ filename "profile-delete-system-folder"

[aceRage]

Status update (Sep 20): marked ready for review. Branch is up to date with main, mergeable (clean), and "Check Documentation" CI is green.

Catch2 results on this branch (Windows, Release):

• [NoSparseLayers]: 44 assertions / 8 cases ✓
• [WipeTower]: 67 assertions / 14 cases ✓

Remaining gate before merge: real multi-filament no-sparse slice to confirm descend-over-tower behavior and change_filament_gcode at layer Z (see accept criteria in the PR body). A combined test build (this PR + #41 + main) is installed in the maintainer's live-testing folder.