The edit-compile-test cycle in 1986 on a large mainframe was approximately 6 hours. A laser printer cut that down to 3.5 hours and that was enough to be able to do twice as much in a day. You can't imagine how surreal the difference was between my productivity at home on my micro computer vs what I got paid for.
The only way to get anything done at all was to have the program more or less complete in your head otherwise you'd be busy for months on a simple problem. Exploratory programming? Forget about it...
The only way to get anything done at all was to have the program more or less complete in your head otherwise you'd be busy for months on a simple problem. Exploratory programming? Forget about it...