I come from CAD and I am having a hard time wrapping my head around how people maintain models in Blender and other mesh tools. If I want, say, a hole through my model, I seems that I have to commit to the general location of that hole, since it is going to be baked into the topology if the model. If I ever want to change the location of the hole it can be a huge job. It gets worse if I have, say, 30 holes in a row, and want to change it to 35 holes in a row with the same total span width.
I can always use modifiers, but eventually I have to apply them and fix all the edge flows they break. If I need to change something at a later date, I will have to redo the edge flows again. This is especially broken if I want to use subdivision surface modifiers, since I will either need detailed control geometry around my hole’s sharp corners, or I will need to do the Boolean after applying the subdivide and be stuck with a lot more geometry to retopo.
How do people deal with this? I make thousands of retrospective changes to a model as I move through a CAD design process. Is 3D modeling more like painting, where you commit to the structure of the workpiece as you go? Are there any tools that I could use to mitigate these issues? So far, I cannot see the advantage of working directly in meshes besides the output being easily UV-mapped. Because of this it is surprising to me that we don’t see more modeling done in parasolid software and then meshed into UV-friendly topologies using automated tools. Either this is just mathematically very difficult to do, or there are some strong advantages to the mesh workflow that I am not seeing.