All pull requests

Pull requests

• fix(blend): rounds meeting at a knife-edge tip are refused by name, not with a missing corner

  #2694·bherbruck opened Nov 4, 2026Collaborator·

  ·
  3

• difftest(step): report bull.stp and moon_buggy.stp body 0 as CPP-ND (run-dependent C++ results, ND-1)

  #2693·bherbruck opened Nov 4, 2026Collaborator·

  ·
  4

• fix(blend): a failing fillet notes why, instead of failing with code 0 and no comment

  #2692·bherbruck opened Nov 4, 2026Collaborator·

  ·
  4

• perf(check): skip B-spline curve analyses that cannot report, and keep their results in the Kernel

  #2691·bherbruck opened Nov 4, 2026Collaborator·

  ·
  1

  3

• fix(surface): iso circle planes: near-parallel planes are never coincident by the helper's own tolerance; on by default (WR-739)

  #2689·ablasik opened Nov 4, 2026Contributor·

  ·
  3

• solidcore-ops: queries, booleans, stitch and merge, on record order

  #2688·jyviko opened Nov 4, 2026Collaborator·

  ·
  1

• [WIP] cleanup: delete the unused solver entry points of fillet_geom (p1, p2, p5)

  #2685·MarkusReichart opened Nov 4, 2026Contributor·

  ·
  1

• fix: a run-on the edges' curves cannot decide keeps the smooth-vertex cap (#2679)

  #2680·brett19 opened Nov 4, 2026Collaborator·

  ·
  1

  5

• [WIP] test: WR-695/696 cases graded against closed forms, not the fork's build

  #2678·MarkusReichart opened Nov 4, 2026Contributor·

  ·
  5

• fix: preserve nested definitions through rollback snapshots

  #2672·altvnk opened Nov 4, 2026Contributor·

  ·
  2

• fix: bound geometry conversion nesting in both directions

  #2671·altvnk opened Nov 4, 2026Contributor·

  ·
  1

  2

• Law corners: the round built as the union of its balls where its radius law dips or peaks (BuildLawCorners, #430)

  #2662·Karl-VonRichter opened Nov 4, 2026Contributor·

  ·
  3

• fix: a rail tangent to a face's edge no longer fails the round at simple_coincidence_checker (#2652)

  #2659·brett19 opened Nov 4, 2026Collaborator·

  ·
  1

  2

• blend: the blend corner is always on; delete the old corner code it replaced (#2400)

  #2653·jyviko opened Nov 4, 2026Collaborator·

  ·
  1

  7

• fix: a round on a round is certified on its inner definition piece by piece, whatever its net (#2570)

  #2650·jyviko opened Nov 4, 2026Collaborator·

  ·
  1

  3

• perf: band_cuts tests only geometry within reach of the ball (#2648)

  #2649·brett19 opened Nov 4, 2026Collaborator·

  ·
  1

  1

• STEP: the product structure as plain data in solidcore-record (#2578, slice 2e)

  #2645·kaltenbrunnermarkus opened Nov 4, 2026Contributor·

  ·
  2

• STEP: the topology entities as plain data in solidcore-record (#2578, slice 2d)

  #2644·kaltenbrunnermarkus opened Nov 4, 2026Contributor·

  ·
  2

• A ball patch read from the B-rep record is certified on its sphere, with one on-sphere certificate (#2357)

  #2641·jyviko opened Nov 4, 2026Collaborator·

  ·
  5

• Chamfers just short of the corner rounds build, or are refused naming the size (#2546)

  #2639·jyviko opened Nov 4, 2026Collaborator·

  ·
  1

  3

• fix: a failed evaluation is not a value in the minimum-radius search (#2574)

  #2637·MarkusReichart opened Nov 4, 2026Contributor·

  ·
  15

• fix: the ellipse check allows the rounding a near-degenerate ellipse's evaluation carries (WR-741, #2539)

  #2618·H1D opened Nov 3, 2026Contributor·

  ·
  1

  3

• fix: preserve seams and surface orientation through local offsets

  #2613·bendavis opened Nov 3, 2026·

  ·
  1

• STEP: Part 42 surfaces and pcurves as plain data in solidcore-record (#2578, slice 2c)

  #2612·kaltenbrunnermarkus opened Nov 3, 2026Contributor·

  ·
  4

• STEP: Part 42 curves as plain data in solidcore-record (#2578, slice 2b)

  #2607·kaltenbrunnermarkus opened Nov 3, 2026Contributor·

  ·
  4