Typical operation
To calculate a de Bruijn diagram, and particularly, to set up some examples whose evolution can be checked by the CAM , the following steps could be followed.
First, select the rule, and be sure that it is installed both in mogrul and the CAM lookup tables. It is better to do this explicitly rather than relying on one of the REC demonstrations to leave the tables behind. The INSERT option within the f1 options which allow rule editing will usually load the tables correctly.
The main menu option d affords the only access to de Moore 
de Bruijn diagrams (access to Moore 
diagrams resides in the menu for that option, but it is much more
limited and requires far less space); the scarcity of data space for CAMEX precludes
having any additional data structures present.
Within the de Bruijn submenu, the class of shift should be selected at once; otherwise the connectivity matrix will be uninitialized, making further results meaningless.
The general survey, matrix powers, or a second level diagram can now be selected.
Having generated a second level matrix, it can be saved on the disk, or used to generate samples on the screens. Only one matrix can be saved per session, the latest if there were several attempts. Samples can be generated and run indefinitely.
Screen samples should confirm whatever type of evolution was selected, but the inevitable discontinuity at the edges of the screen will usually erode the image, starting at the lines where the periodicity fails.
It is possible to begin over again at any time by going back to select a shift class. Different widths within the same class can also be chosen, without having to go all the way back to the shift class options.
The rule table remains intact when returning to the main menu, but all the connection matrices will have to be reestablished when reentering d later on.
Whether the rule has changed in the interim depends on whether mogrul has been altered during that time; von Neumann options leave it intact, but having used INSERT during any Moore option will have surely resulted in change.
Harold V. McIntosh
E-mail:mcintosh@servidor.unam.mx