Abstract
We construct a cellular automaton (CA) with a sofic and mixing limit set and then construct a stable CA with the same limit set, showing there exist subshifts that can be limit sets of both stable and unstable CAs, answering a question raised by A. Maass.