I thought I wouldn’t blog any more about the CRU emails story, but this one is very close to my heart, so I can’t pass it up. Brian Angliss, over at Scholars and Rogues, has written an excellent piece on the lack of context in the stolen emails, and the reliability of any conclusions that might be based on them. To support his analysis, he quotes extensively from the paper “the Secret Life of Bugs” by Jorge Aranda and Gena Venolia from last year’s ICSE, in which they convincingly demonstrated that electronic records of discussions about software bugs are frequently unreliable, and that there is a big difference between the recorded discussions and what you find when you actually track down the participants and ask them directly.
BTW Jorge will be defending his PhD thesis in a couple of weeks, and it’s full of interesting ideas about how software teams develop a shared understanding of the software they develop, and the implications that this has on team organisation. I’ll be mining it for ideas to explore in my own studies of climate modellers later this year…
Pingback: The CRU emails and the secret life of bugs | Catenary
Thanks for the kind mention, Steve, and thanks for putting me on to Jorge’s excellent paper – it totally changed my approach to the piece.
Just wanted to let you know that a companion piece to this one (based on the same interviews but not Jorge’s research) has gone up at S&R here.