Maintainers have been backporting some of the fixes to the v1.1 branch where possible — 19 backports in the past 7 days. The list of changes in this recap applies to the main development branch (future v1.2).
This week in FreeCAD development:
Draft:
- YashSuthar983 added snapping to knots of BSplines (PR#26571).
- Roy-043 fixed the autogroup behavior when the active group is a layer (PR#27102).
Sketcher:
- PaddleStroke fixed the issue where dimensional constraints would visually match the lines when using on-view parameters (PR#25848). He also added a tooltip that displays the entire expression for constraints with an expression (PR#25954) and fixed disappearing construction lines during drawing (PR#27195).
- vcheckin fixed an intermittent crash on sketch exit (PR#27129).
Part and PartDesign:
- kadet1090 fixed two regressions in the MultiCommon boolean operation (PR#27012).
- theosib fixed mirror() regression with non-identity Placement (PR#26963).
Assembly: PaddleStroke contributed multiple fixes and improvements:
- Fixed the slider joint of the sub-assembly not working when inserted as a grounded part (PR#27206).
- Fixed isolate not working on sub-assembly components (PR#27173).
- Fixed assembly activation issues (PR#27194).
- Removed an extra solver message when the file is closed (PR#27210.
- Added support for single click to open exploded views in read-only mode, without entering the edit mode (PR#25456).
TechDraw:
- Lgt2x fixed Qt5 compatibility (PR#27253).
- WandererFan fixed the default circle centerline line style (PR#27134) and fixed a release blocker where the projection group jumps to the origin (PR#27094).
CAM:
- Daniel-Khodabakhsh fixed an error that would occur when Start depth equals to Final depth and Safe height (PR#27258).
- Luvtofish fixed and updated the Dyna_4060 postprocessor (PR#27202).
BIM/Arch:
- Roy-043 had eight of his pull requests merged, including these: fix for BuildingPart area calculation for indirect children (PR#24848), support for relative paths in hybrid IFC files (PR#24190), fix for subtractions not working with spaces (PR#27239), a test for horizontal area of tilted cylinders (PR#27108).
- Arusekk fixed a bug where an ArchBuildingPart would not move the child object base and only move the child itself (PR#27237).
- furgo16 added a task panel for wall options so that you can edit, e.g., alignment and width in one go (PR#26758). He also fixed a bug where a baseless wall with additions would have a different icon in the Tree View (PR#27277).

Other changes:
- pieterhijma contributed numerous improvements to the documentation of C++ API, expressions, and element mapping (PR#25194, PR#25195, PR#25196, PR#25197, PR#25198, PR#25199).
- kadet1090 added a TimeTracker utility that can be used to orchestrate code with time measurements in an easy way and detect performance issues (PR#26305).
- nishendra3 fixed angle measurement in the Measure tool (PR#27254).
Roy-043, graelo, oursland, chennes, phaseloop, Krrish777, timpieces, YashSuthar983, cnaples79, leoheck, kkocdko, greg19, mosfet80, furgo16, xtemp09, PaddleStroke, captain0xff, 3x380V, pyro9, and Lgt2x contributed additional improvements and fixes.
If you are interested in testing the latest weekly build, you can grab it here.
PR stats: since the previous report, 88 pull requests have been merged (including backports to the v1.1 branch), and 29 new pull requests have been opened.
Issue stats: overall, there are 3199 open issues in the tracker, up by 4 from last week. There are 4 release blockers for v1.1 currently, down by 3 from last week.
2 responses to “WIP Wednesday, 4 February 2026”
The Fillet tool shows inconsistent behavior.
Problem description
When applying a fillet simultaneously to two edges, an inconsistent behavior can be observed depending on the geometric orientation of those edges relative to their common tangency edge.
Case 1: Fillet on Edge 1 and Edge 2 (common edge: Edge 4)
When the fillet radius is smaller than half the distance between the two edges, the operation succeeds correctly.
When the radius is equal to or greater than half of that distance, FreeCAD reports the error:
Input error BRep_API: command not done
This behavior is geometrically correct.
Case 2: Fillet on Edge 2 and Edge 3 (common edge: Edge 5)
When the fillet radius is greater than half the distance between the two edges, the operation is executed without any error, but the resulting geometry is invalid, producing an irregular solid.
The difference between both cases is that:
In the first case, the centers of the fillet radii lie on the same side of the common tangency edge.
In the second case, the centers of the fillet radii lie on opposite sides of the common tangency edge.
In this scenario, the fillet tool should fail or report an error instead of generating an invalid geometry.
Additional note:
The error “Input error BRep_API: command not done” should only be raised when the fillet radius is strictly greater than half the distance between the edges.
When the radius is exactly equal, both fillets share a common tangency point, which is geometrically valid.
See attached screenshot
https://imgur.com/a/QUWWZnN
Best regards,
FreeCAD
Version: 1.2.0dev.20260204 (Git shallow)
Build date: 2026/02/04 19:46:02
Build type: Release
I will keep approving your comments, but please be aware that developers will likely not be reading them. This is not the usual communication channel with the dev team.