genejockey 

When I started at Genentech in 1982, they had a Vax, what kind I don't know, running Unix. We were using CRT "dumb terminals" and doing a lot of hand-annotation of sequences on those big green-and-beige fanfold lineprinter printouts. Everything was command line, because back then everything WAS command line.

We had a programmer who was really into finding all the algorithms people were developing for DNA sequence analysis, and adapt them for our use. You'd say, "Hey, Colin! I could really use a program that does (insert idea here)." and he'd say, "You know, I was just reading about something that might work. Gimme a week."

A week later, he'd give you a slip of paper with the pathname to the new program, and the syntax, and say, "Try it and let me know!" And then as you worked with it and reported back to him he'd refine it till it was what you wanted. Then he'd let others know it was there. By 1991, when I left Genentech, we had a whole suite of proprietary sequence analysis programs, which was WAY better and more user friendly than anything else I used for at least another 5-8 years.

These days I'm working in R on a virtual machine, analyzing sequences on a scale we couldn't even dream of then. And once again, everything is command line. And I'm still a crappy typist.