Careful guys, some of the software I wrote in Fortran (1977-1980) and in COBOL (1981-1989) may still be running, keeping Fortune 500 companies in business all these years, while satisfying customers with accurate information about their accounts. A 45-year track record of success. How much of today's software will be able to say that in 2066?

My point is, that I agree with Mika, longevity is indeed a good measure of a computer language.