Much of the actual day to day work is. Typically graduate students, so they’ll be 22-26. That’s not a critique of their intelligence or potential. Students get progressively more experienced of course, but professors aren’t writing code most of the time.
A problem with academia in general is the lack of staff positions. Post docs finish their time then it’s either leave academia or become a professor. There’s few positions for those who want to just do research as a career, rather than pushing for a professorship. This means there isn’t a stable and experienced core of people.
Obviously slanted to certain areas (OSes and languages, rather than say word processors), relevant to research, but still.
It has not historically quite important.
Of course, it would be great to fund experienced people just to do this - and a better use of the money currently subsidising commercial R & D at the moment in many countries.
Yeah, most of it is. I got a look behind the curtain when when my son got a master's. His PI was wrote a huge Python program then left and he inherited it. The new PI is completely clueless. They all have other, more pressing things to do, instead of doing proper software engineering.
When I was at UCB in the 80's, a lot of incredible things happened (Berkeley UNIX), but they had a LOT of staff members that did a lot of the work. And that had PhD students (Bill Joy, Sam Leffler) who were insanely smart and spent most of their time doing proper engineering on their projects. And, btw, I was one of those staff members. I saw all aspects of it, because the project I was on was used by a lot of people in the CS dept.
I wasn't actually criticizing anyone. I think it's just the way it is.
If it was a primary function and was staffed independently of educational programs, it could work and be a great teaching tool for actual students.