Snapping when initially drawing an insert mesh object is sorely needed. Holding the shift key does uniform scale instead. One workaround to snapping is to turn on symmetry (not radial) and draw your object where the two dots converge, but this is only useful if you need something along the dead center of either the x, y, or z plane. Holding the space bar to reposition the inserted mesh object will sometimes crash ZBrush when doing this though (InsertCylinder particularly), and uniform scale via the shift key doesn’t even work with customized insert mesh objects, only the predefined ones (cylinder, sphere, cube, etc). Last, but not least, is the ability to smoothly subdivide an insert mesh object after drawing it onto your model. Subdividing results in facets with the way it works right now.
Another idea I had was a way to automatically draw an action line, say centered to the current object/subobject/insertmesh and perpendicular to either the x, y, or z plane. A button similar to the current floor button would do. Or a way to quickly draw an action line on the canvas with a hotkey to center it to the current object. The point is to speed up the process when doing a lot of action line related tasks. Constantly having to manually redraw and double check it for basic tasks is cumbersome, and creating a straight perpendicular line can be tough with small objects, especially if they have no flat surfaces.
Better/easier placement when drawing curves would be helpful. For example, locking the start and end points to where you click on the surface, as well as drawing the curve based on the direction of the normal or x/y/z plane rather than the view. This would help when trying to draw a curve in a precise location on your current mesh. Right now it is impossible to draw a curve on the side of a model because the side you wish to draw on will always be out of view thanks to basing the curve on screen alignment rather than a polygons normal. Yes, you can use an action line to fix it afterwards, but trying to do a precise rotate and move is harder than it appears, especially when using radial symmetry. Snapping the curve to the surface when moving it via an action line would be useful too, and not just when first drawing it as mentioned above. Imagine a cube sliding along the surface of a sphere, like two magnets, for visualization of what I’m trying to get across with this one.
The ability to change the default Lightbox directories would be a welcome addition (as would a working preference button for Lightbox lol). I assume that like many others, I do not use the default C: drive (which is just a small 60GB SSD on my PC) to store my work. Lightbox numbering needs to be fixed as well. For example, Shadowbox256.ztl should not come before Shadowbox64.ztl (Windows gets it right though). Pixologic should seriously consider adding 7z or some other compression to their file formats as well, especially files in the project format which tend to get very large, which is compounded by incremental saves. The difference in file size between compressed and uncompressed was pretty huge, and will help to save a ton of hard drive space in the long run.
The ability to reposition a DragRect+Alpha using the space bar would be very useful, as well as the ability to rotate it with or without snapping at the same time. Alphas are a simpler way to add detail versus creating custom InsertMesh brushes.
A ghost mode that stays on when drawing/adjusting the placement of a boolean objects on a dynamesh, as well as the ability to invert which object is ghosted on the fly. This is just a minor change to the current version of transparency and should be easy to change. A delete button for boolean objects next to the add/sub/and buttons could be useful, primarily because you can run out of undos. Would it be possible to create an unlimited undo system, perhaps using the hard drive instead of memory? That would also fix this potential problem. I’d also like to see those three buttons (add/sub/and) work properly, or be able to change from Zadd to Zsub (or vice versa) after drawing a boolean object. It is really annoying to put a huge amount of effort into getting the size, shape, and placement just right, only to find you forgot to hold the alt key when drawing the boolean object.
Hmm, how about an automated incremental save system? Or just a simple plus button in the save as dialog (like in 3ds Max) would work.
That is all I’ve come up with for now based on my Z4R2 testing. I’m sure I’ll find others as time progresses. Hopefully I didn’t repeat anything that was already on the list. 