Torsten Paul

The example looks reasonable and useful. I think we probably need to have warning messages for cases that are not supported and a clear definition of the extended support.

I'm not convinced it should work that way. Maybe there are other options to handle that case, but showing tabs without the editor looks wrong to me.

Yes, I mean more logically, or maybe even more specific workflow wise. Hiding the editor and then still using it partially likely opens a can of worms. So I wonder...

What would be the reasoning for ignoring the view box?

That will potentially expose random transformations from the graphics program. I'm not sure that's a useful thing to have. I feel that's a big source of problems with random files....

Hmm, so it would be more a `clip = true/false`.

@DavidPhillipOster could you put that in a separate issue please, that seems a bit unrelated from the scope of this issue. The path looks certainly a bit funny `d="M4 4.5h7m4...

Similar to #722 I think having an "export()" function that just does the export at the point when it's evaluated, is not matching the OpenSCAD logic at _describing_ an model....

I would avise against using this `-D` behavior, it's defined for variable assignment but currently can be used to pretty much anything else too. Instead you can use something like,...

It would help if there's also a recovered version of the file, assuming that works as mentioned in the screenshot.