We could introduce a "restore-mode" (demo-mode) for computation, in which every computation restores all settings, under which it was previously evaluated, before it is re-evaluated. Could be nice for demos, but should not be standard because it slows down computation.
We could introduce a "restore-mode" (demo-mode) for computation, in which every computation restores all settings, under which it was previously evaluated, before it is re-evaluated. Could be nice for demos, but should not be standard because it slows down computation.