I encountered a funny comment in the "joke of the day" file while working as a UNIX sysadmin at PacBell. It was similar to "Those who can, do. Those who can't, teach. Those who can't teach, write. Those who can't write work for the Bell Labs Record". This is from a time in the early 1980's when UNIX was still very much an inside the Bell System product. I had read and understood the fully commented kernel source from the University of New South Wales in the Summer of 1981. It was June of 1983 when I finally laid my hands on a 3B20 at PacBell. Stroustrup's early C++ was on the machine. Life was good. C++ was more fun that writing raw assembly language on the 8080, Z80, 6800 and 6809. Writing devices drivers for UNIX was duck soup after 4 years of doing embedded microprocessors. By 1985 I was collaborating with Phil Karn, KA9Q, to do TCP/IP over AX.25 and applying the TCP/IP skills to upgrading PacBell's networks. I was porting David Korn's shell to 30 different platforms and pleased as punch that he was very responsive to bug reports. No complaints about my career path.