Dimensioning should be improved by git commit 41bce2eaf4.

JarvisOac wrote: ↑Sun Apr 14, 2024 8:55 pm

You are missing the sleep() between recompute() and find_vertices().

user1234 wrote: ↑Fri Apr 12, 2024 6:04 pm The dimension of the first side is wrong. See real measurement,

Yeah, it is the real value + (removed area of the break - gap width). If have a >= somewhere that should be a >.

JarvisOac wrote: ↑Fri Apr 12, 2024 9:00 pm

Is the code you posted not working after 10 seconds? Or is that just a minimal example? 1 second is enough here for the sample code.

If there is a more complex version, I'd be happy to look at it.

Version 0.1 of Broken View is available after git commit 74556bced8

SLJeans wrote: ↑Fri Apr 12, 2024 1:40 pm

user1234's suggest is the "correct" fix. For a quick and dirty workaround, use the Perspective projector with a long focal distance instead of the standard projector. Perspective sometimes works when the default does not.

The HLR & face finding threads started as a result of doc.recompute() are not finished before the call to getVertexBySelection. Until the HLR thread completes, there is no geometry to provide a vertex. If you add a short wait it should work. The wait status methods are not exposed to Python, so...

This is really odd. getVertexBySelection works from the console, but not from a macro. Will have to dig a bit on this one. >>> # doc = App.getDocument("touchedOnClickTest") >>> # obj = doc.getObject("View") >>> ### End command Std_SendToPythonConsole >>> obj <DrawViewPart object...

Mejico wrote: ↑Fri Apr 12, 2024 12:49 am

Any chance of a sample file showing the problem? It's hard to debug pictures.

user1234 wrote: ↑Tue Apr 09, 2024 10:20 am Or had not the time yet.

This is the winning answer.