Sarah Mei tweeted this in response to news that researchers at Rice were getting $11 million for a souped-up autocomplete:
These Rice University folks do realize that "coding" is about 5% of what software developers do. Right? http://www.engadget.com/2014/11/09/darpa-pliny-coding/
Hope they can optimize the shit out of that 5% with their 11 MILLION dollars. That is surely a good use of my taxes.
Academic research on software engineering has been a decade behind current practice since I've been practicing. Some contributing factors:
1) The folks with PhDs often haven't been practitioners for years (if ever). They don't know what state of the art looks like.
2) They often study students, so their results seem curiously divorced from the practices of a typical team with a mix of experience levels.
3) When they don't study students, they study large companies. Large companies are typically a decade behind the leading edge on process.
4) The researchers lack the network to study small companies, and small companies themselves are often less interested in being studied.
5) Many researchers don't think startups are worth studying. They're "not doing real engineering," or "too small-scale to be interesting."
I don't know how to fix this. Some parts of computer science are way out ahead of practice, and I wish this were one of them.