Agreed with Eld - I don’t think there’s any case where you would want to apply EdgeSplit. And on the other side of that, EdgeSplit (with From Edge Angle off) doesn’t change your mesh unless you tell it to by marking stuff, so there’s no downside to having it on all the time.
The confusing part is that Set Smooth/Set Solid often does the same thing as EdgeSplit.
IMO, it would be easier if From Marked As Sharp would always be on, with the modifier just handling From Edge Angle.
I personally wouldn’t want the extra processing overhead involved in doing edge split unless I asked Blender to do so. Being able to turn it off also makes it possible to work with very large meshes in Edit-Mode that would be very difficult to handle if it was always on (think hundreds of thousands of triangles). These same objects however are fine to be split in object mode because your not constantly updating the cached split mesh.
Cheers,
Briggs
…hadn’t thought of that, but maybe that explains why blender and lightwave are really quick and responsive with loads of objects and polygons and maya is sluggish and un-responsive on the same scene…
previously i’d thought it wouldn’t be a big deal as you can quite happily display flat shaded (where every single edge is sharp) without any real performance cost…(except sculpt mode where partial re-draw forces smooth shading for speed)
Could you explain why that’s so much of a slowdown? I’m not familiar with Blender’s code, but in theory, Set Solid is the same as EdgeSplitting every edge, and I don’t notice any speed difference between Set Smooth and Set Solid. (In fact, I just did a test and I don’t even notice a difference between EdgeSplit and no EdgeSplit).
It could be because there is shading techniques for solid shading that have been around for a very long time (since polygons were first being used). In fact solid shading existed before smooth shading did in the early days of CG as there’s no interpolation between the lighting values of different faces.
I love blender’s edgesplit modifier but performance wise it is ridiculously slower than max and maya’s solutions. It doesn’t take much for this modifier to really bog down a scene.
Set solid only tells Blender to draw every face-vert in the mesh with the face normal instead of vertex normals. This is quick and easy to do.
Edge split as it is now requires a full copy of the current mesh with split edges to be built. To do this full adjacency data must be recovered as well, which in of itself is not slow if you only do it once in a while. However doing this every redraw is not going to be fast.
Torq: The Edge-Split modifier in B-Mesh branch is supposed to be much faster, and can be made faster still by caching and incremental update similar to what Sub-Surf does now (although that will require more memory).
All optimizations aside this belongs in the modifier stack since its built for this sort of thing IMO.
Cheers,
Briggs
Thanks, Briggs, that makes sense. It’s a bit of a hassle to add the modifier to tens of objects, but I guess it’s the better option here 
.
You can just add it to one and use the copy menu (CTRL+c) to copy it to the others.
it should be the same as setsmooth/set hard, not a modifier.