The obvious answer is that the statement (as underlined above,) is not true.
It is not true no matter how many sages say it, because it is worded incorrectly.
The true “rule of thumb” is that: “All PRIMITIVES should be on Layer0.”
There are times when changing the active layer is desirable. Such as inserting a series of component instances on a specific display layer. It is much easier and faster than selecting them all later and changing their associated layer via the EntityInfo inspector.
But this protocol usually involves the conscious change of active layer, the insertions, and the switching back to
It is not the software’s fault if the user doesn’t use it correctly.
Dumbing down the software, and taking power away from users who DO use the software correctly, is not the correct answer.
Besides group and component instances, … dimensions and guide elements (lines and points) can be put on display layers (ie, layers that are not
Things to be wary of:
Images. Images are a special sub-kind of component instance until exploded. If you explode them, take care to move any primitives (edges & faces,) to
Intersecting: Some intersect or boolean tools will put the results on the active layer. These results may be primitives. It can be powerful and an advantage to temporarily have these results on a layer other than
Layer0. But if they are left in the model after processing, they should be moved to
Layer0. (Plugins that use intersect, usually do this, if written well.)