Never mind the quality, look at the results
Climate scientists should not write their own software, says researcher | Environment | guardian.co.uk
A study by a computer scientist at the University of Toronto suggests that the computer models used to predict climate change may be undermined due to a lack of programming expertise.
Steve Easterbrook at the University's Department of Computer Science, has had his paper, Climate Change: A Grand Software Challenge, accepted by the 2010 FSE/SDP Workshop on the Future of Software Engineering Research. In the paper, he suggests that because many climate prediction software modelling tools are built by climate scientists rather than software engineers some of the resulting software has room for improvement.
Prof Steve Easterbrook then comments ...
Thanks for writing about my research. The article itself is quite good, but your headline and the opening paragraph are entirely wrong - my research shows the opposite is true. The fact that the scientists write their own code is one of the most important success factors in ensuring global circulation models produce high quality, scientifically valid simulations.
Please change the headline as soon as possible. I'd be happy to provide you with a more detailed interview if you would like.
Prof Steve Easterbrook
I think he means that because some of it is well written and all of it gets the right answer it doesn't matter that some of the code is crap. It is bit like going to church, some of the hymns are well crafted and sometimes trendy vicars insist on strumming out their own compositions on the guitar, it doesn't matter because it all leads us to salvation.