Programming in 2010? A scientific and industrial challenge

The development of software technology in the industrial world is rather discouraging for researchers today. Languages and systems offered by software houses and computer vendors are still very poor if compared to the state-of-the-art in advanced research. Fortran and Cobol are still among the most widely used programming languages, relational databases are just starting to penetrate the market. The breathtaking speed of transfer of new hardware technology from research labs into every day life is not at all matched by a comparable development in the software area. It is very hard to explain this enormous discrepancy of acceptance between the hardware and the software domain. Is it a lack of flexibility on the user side, or a lack of insight in the importance of software by the manufacturers, or is it just a matter of complexity? Is a new piece of software really that much more complex than a new hardware component?