Comment by thesz

First, surrounding your code is other code that very probably have a clue on what you should do. You do not have to go at the definition to see what to do.

Second, for one written/changed line I have read, perhaps, couple of hundredths of lines of code. This results not in a "fBeggy is a field of a class CBaggy" kind of understanding, but "the general approach to solve problem like this one at hand will be such (see this part) and such (also see this part)."

As a software support engineer, I have twice as low yearly output of (real) LOC (11.5K/y) than is expected from software engineers at, say, IBM. Mostly from the need to read substantial amount of code.