smlib: solidcore's five new fork-carried opt-ins are on; the trim end-span opt-in waits

fillet_chamfer_cone_stored_as_cone (WR-572, smlib@b5743ec), offset_point_vertex_caps (WR-531,
smlib@d5dadaa), approximate_surface_bezier_starter (WR-573, smlib@eae96cb),
proto_edge_splits_at_captured_vertices (WR-579, smlib@9eec159) and stitch_solid_nests_cavities
(WR-569, smlib@351c57c) are on: their fork commits are in SMLIB_REV 985faaee.
bspline_trim_keeps_rational_end_spans (WR-575, solidcore#1711) has no fork commit, so it goes on
HELD_BACK.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MNXLXkZLQ6T7TSi4wswcmk
This commit is contained in:
Brett Lawson
2026-10-30 18:19:03 +01:00
parent 9c1f8904e0
commit 63864b2984
+8 -2
View File
@@ -186,14 +186,20 @@ pub(crate) fn fixes() -> UpstreamFixes {
f.fillet_knife_edge_rounds = true;
f.taper_wall_closes_at_apex = true;
f.revolution_round_offset_collapses = true;
f.fillet_chamfer_cone_stored_as_cone = true;
f.offset_point_vertex_caps = true;
f.approximate_surface_bezier_starter = true;
f.proto_edge_splits_at_captured_vertices = true;
f.stitch_solid_nests_cavities = true;
f
}
/// Opt-in fixes left off because the C++ fork (`SMLIB_REV`) does not carry them yet, so turning
/// them on would part this backend from the C++ one. `precise_properties_trim_base_v`
/// (solidcore#1664, WR-563) has no fork commit.
/// (solidcore#1664, WR-563) and `bspline_trim_keeps_rational_end_spans` (solidcore#1711, WR-575)
/// have no fork commit.
#[cfg(test)]
const HELD_BACK: &[&str] = &["precise_properties_trim_base_v"];
const HELD_BACK: &[&str] = &["precise_properties_trim_base_v", "bspline_trim_keeps_rational_end_spans"];
fn fresh() -> Kernel {
let mut k = Kernel::new(&KernelOptions::default());