Similar to what @papyblaise suggested, you can sweep a triangle around the complete outline of the letter. I have used Part_Sweep. But the same principle should work in the PartDesign WB. 1) Use Part extrude with a taper angle That seems to be problematic. My guess is due to the splines in the outli...

I think that it is unlikely that this is the Draft grid. For one thing it does not have the default appearance and colors of a Draft grid. The OP should post a file.

The trick removes the duplicate entries in the tree, there were never duplicate objects.

Angular constraints are often advised to prevent flipping.

1. Crash Behaviour Without a file it is hard to help you. It is not an issue that occurs with all BuildingParts. 2. Multiple Instance Object Behaviour Confirmed. You can fix this by dragging the rectangle and the structure directly nested in the BP out of the BP and dropping them on the top (docume...

Draft has several geometry functions: import draftgeoutils draftgeoutils.geometry.findDistance(App.Vector(), [App.Vector(1,0,0), App.Vector(0,1,0)]) edge = App.ActiveDocument.Line.Shape.Edges[0] draftgeoutils.geometry.findDistance(App.Vector(), edge) To do this manually you can use the Draft_Dimensi...

This is complex. For one thing the legacy importer obeys the Draft internal precision setting: Edit > Preferences... > Draft > General settings > Internal precision level You may want to crank that up from the default value of 6. You are correct in saying that the line segment is exactly 1.35 drawin...

Thanks for those links. I'll try to look at them, but this stuff may prove to be over my head. Keep in mind that the current algorithm supports gables, and roof segments with different angles based on consecutive parallel edges. Also the roof segments need not connect at the top. In other words: the...

There is something wrong with the DXF. The MEASUREMENT is 0 (zero), which indicates imperial units. The legacy importer ignores this setting but the new importer does not. But obviously the new importer does not get it right. It uses a wrong factor (1/25.4 instead of 25.4) PR: https://github.com/Fre...