The document discusses the Compact Classifier System (CCS), which uses a compact genetic algorithm (cGA) and fitness function to evolve rule sets composed only of maximally accurate and general rules for classification problems. The CCS is tested on multiplexer problems of varying sizes. It is shown that CCS succeeds in evolving the optimal rule sets for smaller problems but performance decreases with increasing problem size, demonstrating issues with scalability. The binary rule encoding used is found to produce a large number of unmatchable rules which negatively impacts scalability for larger problems. Alternative rule representations are suggested to improve scalability.