Peter Percival wrote: > Arturo Magidin wrote: > >> The theorem is **NOT** about "patterns". The theorem does **NOT** say >> that there is an algorithm for producing a coloring, as you seem to >> think. The theorem says that **GIVEN** a map, a coloring exists *for >> that map*. > > May I ask if there is a practical algorithm? Clearly there is an > algorithm: colour the given (finite) map in all possible ways with four > or fewer colours, examine each colouring in turn until an admissible one > is found. The theorem ensures that one will be found. But that would > be quite impractical.