I just uploaded another version of the C file, this time with a copy-paste-adjust-warnings block from the JS version, and it worked right away.

For the record, you have to press ENTER to compute the next iteration. Add to this the cost of displaying data through the Xterm and you understand that it's far from the true speed that this algo can reach. But I'll optimise speed for #AMBAP: A Modest Bitslice Architecture Proposal later.

For now, I rejoice that it worked just like I expected :-)

                       #
                     ####
            ##     #### ##         ##
          #  #   #   ## ###        ##
 ##      #       #   ## ##
 ##      #      #   #####
         #       ###   #
          #  #
            ##

                         #
                          ##
                         ##

              ####
             # ## #
             ##  ##
             ##  ##
             # ## #             # #
              ####               ##
                                 #