The use of the parameters k and i as size_t seems unnecessary. We can change it to int and change the System::addConstraintInternalAlignmentKnotPoint function accordingly. All variables involved are actually int . Edit: actually a cast is needed there too. They are used as, but i and k are also ind...

There's a toponaming and linkstage3 subforum. Maybe someone will move this thread there. Alternatively, you could download a more classic version and attempt to reproduce the issue.

chrisb wrote: ↑Fri Nov 18, 2022 5:31 pm I will not prove it here, because it is always possible to create an ellipse which has the same offset of minor and major diameter. I use circles of equal diameter to show the distance:...

I see. So basically you've just given a counterexample . Thanks.

chrisb wrote: ↑Fri Nov 18, 2022 1:25 pm It should be noted, that the offset of an ellipse is no ellipse.

Do you happen to know literature where this is proven? I had been looking for something similar for B-splines and a similar result was shown in the abstract but I wasn't able to even see what reference was used there.

My stance is as follows: ... 2) Changing the algorithm needs to be done before release for sure. At the same time, I think it would help not to change it immediately. The current version of the algorithm has been deployed for some time and there is not a single report of a hang (I have warned about...

There is a while (1) loop in GCS::System::identifyConflictingRedundantConstraints() . There is a possibility that this does not exit leading to FreeCAD hanging. This thread is to discuss what can be done about it. Originally reported here: https://forum.freecadweb.org/viewtopic.php?p=639280#p639280....

The current best workflow for stl export is to mesh the shape in the mesh workbench adjusting parameters as needed and then export that but this is cumbersome, specially if the shape is modified and you need to re mesh to re export. That can still be automated, I believe. Especially with FC Nodes i...

You must start it from within a terminal. At program exit it prints a list of not freed memory and ideally tells you where this memory was allocated. I see. Running the whole test it just gets killed. Running a single instance and closing I get this: Tracer caught signal 11: addr=0x18000592f8 pc=0x...

wmayer wrote: ↑Thu Nov 17, 2022 12:23 pm https://forum.freecadweb.org/viewtopic. ... 08#p480408

Thanks. I tried this but nothing changed. Is there any way to know if asan is active? I don't see any difference in the output.

same, however the system monitor lists freecad as taking at most 1.8Gb I believe it's the same for me. Something (I guess my OS) kills the test and that only gives back about 4-8 GB The number is well above 16-20 GB for me... 16-20 GB is insane. Do you run the tests 10 times or is it the total memo...