Abstract :
The origin of the year 2000 problem is outlined. Its possible forms-data rejection, wrong results, and system failure-are listed. A lack of documentation in early programs is noted. Costs of correction are discussed. The possibility of using assembler code to correct COBOL programs by using spare bits in the year byte is discussed. Insurance against effects of the problem is considered
