View Issue Details

IDProjectCategoryView StatusLast Update
0000950FreeCADBugpublic2014-09-22 13:35
Reporternovalis_dt Assigned To 
PrioritynormalSeveritymajorReproducibilityalways
Status closedResolutionunable to reproduce 
Summary0000950: Shared solids in CSG cause strange behavior
DescriptionI'm using the FreeCAD daily build from the Ubuntu PPA on x64_64. I apologize for having no idea what the correct terms for things are; this is my first time using any CAD software.

To reproduce:
In the part view, create two boxes of standard 10mm size. The second box's position should be (10,10,0) -- that is, it should have an edge coincident with the first box. Now, create a cylinder of radius 2, height 10, position (10,10,0) -- that is, its axis should be coincident with the shared edge. In the project tab, select the first box, then additionally the cylinder, and click the Cut icon. Now, click the little arrow next to the new "Cut" in the project view to open it up. Select the second box, then additionally the same cylinder (that's part of the first cut), and click "Cut".

In other words, I'm trying to use the same cylinder to cut two different boxes.

The problem is that then in the project tab, the second cut (or perhaps the first cut -- it seems to vary) only lists the box, not the cylinder. And the more that I try to use objects inside booleans in other booleans, the weirder FreeCAD's behavior gets (until it crashes).
TagsNo tags attached.
FreeCAD Information

Activities

yorik

2013-01-08 10:52

administrator   ~0002821

Hi,
The tree view showing "nested" items (cylinder inside the boolean, etc) is only cosmetic. You should be totally able to reuse the inner objects in other operations, although several freecad users would argue with you that it is not a best practice.

Actually there is still a debate going on about what the tree view should show in a case like this.

The crashes you experience are probably another problem, very likely due to boolean operations with many "special cases" (coincident edges, etc). The opencascade kernel is a bit weak in that area. If you can isolate a specific case that can be reproduced and always causes a crash, we can try to solve...

novalis_dt

2013-01-09 02:33

reporter   ~0002833

Unfortunately, I can't manage to reproduce the crash (I tried a few different ways). I'll let you know if I ever do figure it out.

Issue History

Date Modified Username Field Change
2013-01-08 07:38 novalis_dt New Issue
2013-01-08 10:52 yorik Note Added: 0002821
2013-01-09 02:33 novalis_dt Note Added: 0002833
2014-09-22 13:35 shoogen Status new => closed
2014-09-22 13:35 shoogen Resolution open => unable to reproduce