Ary Borenszweig:
It isn't just a feeling :)
Researchers are mostly judged and paid (and career advancements are mostly based) on the amount and "impact" of papers written. It's uncommon for "how ideas work in production code" to have a positive influence on career. To this you have to add the fact that if they pay you to study and invent new ideas, and you want to have fun inventing them, you will go look where are the cutest puzzles to solve (like very complex functional tricks).
This is causing a bad and costly disconnect between practice and research in computer science and coding, and this is hurting our society. Even researchers in the private sector like at Microsoft are plagued with a very low ROI because of that.
To solve this problem there is a growing need to tie the career advancements and pay of computer science researchers to the solution of practical problems. I love free basic research, but here there is a growing problem in need to be solved, for the society.
The situation is improving only a little, but this far from enough:
http://wattsupwiththat.com/2012/04/17/the-journal-science-free-the-code/ Bye, bearophile
