sorting 3980

snip

Yes, I'm generalizing from one of my own experiences, and you're right that I did get better, as time went on, at figuring out how to modify-extend-debug other people's code (which is mostly what I was doing). With practice one gets better at what I think is the critical skill -- figuring out what parts of the existing code you must understand and which parts you can ignore in making a given change. Where the five years -- or less, for someone with more practice 1 -- pays off is in knowing the structure and quirks of the existing code.

1 Someone with practice in figuring out other people's code would probably be at the point of being able to do *something* in a matter of -- hours? days? -- but it's hard for me to imagine anyone who wouldn't be significantly more capable after a year.

(Did I mention that this was 100K+ lines of not-well-structured and poorly-documented 15-year-old FORTRAN? Small by comparison to some code bases, but not quite in the "toy" category by the standards of the 1980s?)

Anyway, when the little company folded, a few years after I bailed out to go back to school, many of the programmers had significant problems finding new jobs, and as far as I could tell a key problem was that they had spent their time learning more and more about less and less, and had trouble convincing would-be employers that what they had learned was applicable ....

You are so right that someone with some experience, and previous exposure to something conceptually similar, can quickly learn enough about a new tool to do use it productively. (It's kind of cool when you realize you know enough to do that .... ) I'm having trouble understanding how anyone can develop real fluency in a short time -- seems like most tools have enough details or quirks that are only mastered with time and practice -- but maybe I'm starting to quibble about semantics here.

snip

-- B. L. Mbuttingill ObDisclaimer: I don't speak for my employers; they return the favor.