This is my second idea. If the loading of the environment takes so long with multiple Z-layers, I could replace that reals Z-height for visual tricks like you see on the screenshot. Do you think that is a good solution?
I'm not sure that I get it, but you can try it and let me know how it goes.
Often when I want to try a new process, I build it another cca file to see how it work. Then if all goes well, I bring it back into the actual project cca.
Well... look at the edges of the pool. You'll see that it looks like the water is one level below the grass level. But this is not done by real Z-stacking, but with "visual tricks". I mean the grid that you see over the water are the actual tiles that are placed.
So it's not like grass-Z is 2, water-Z is 1... but it's pure visual.
Big update; my isometric prototype is done and working very well.
- The movements are smooth in any directions. (W,S,A,D,Space)
- The 3D collision is pixel-perfect.
- The layer sorting process is optimised and giving great results. You can move around, over and under objects.
Next, I want to rework the movement and collision loop to minimise the number of loops.
Then I'll replace all those many small cubes with fewer and larger active objects.
I'll also add support for multiple moving objects.
I forgot to mention the extensions: Isometric Grid 2, Value Finder and Leyer Object