You're right Tobias, maybe I phrased that badly, I seem to do that a lot lately.
Its just 'trickier' to manage in my specific implementation.
Im mainly crying over the fact that I cant create a mesh in a malloc'ed
block that I can free at will, using standard 3DGS functions,
that is that ent_clone consumed memory blocks cannot be free'd individually.

But I should mention, as a Bug report, this issue is resolved as NOT a bug.
Even though I wish it were otherwise.