I started working for Honeywell in 1962. This was my first job. I wrote the code generator for a FORTRAN compiler. My great innovation was to read the intermediate representation tape backwards (Honeywell tape drives could do that) because it was faster than rewinding the tape because of the heads having to move away and then reset.
When that was done I became project leader of a FORTRAN compiler project for another set of hardware. Bob Freiburghouse who had newly joined Honeywell was the chief compiler guy.
For both products we wrote our code in assembly language on paper; key punch operators created the card deck. The card decks were sent to a machine in another building and several hours later you got a box of paper - sometimes full of many pages of zeros. It is really a wonder that any projects ever got done. For the second project the OS we ran under was so bad that we wrote it out on tape, loaded our own to run the compiler and the results of the compilations and then reloaded the original operating system when we were done.
When this project was done in early 1966 Bob and I joined the Systems planning group where we tried to lobby for creating a decent operating system, writing programs at terminals, and in a higher level language. We didn't make much progress. In the fall they sent us to an Industrial Liaison symposium at MIT where the subject was Multics. At the end of that we both looked at each other and said let's go work there. We knew a couple of the GE people who had left Honeywell and contacted them. Somewhere around this time GE attempted to cancel the project so there was a delay in the hiring process until MIT persuaded them otherwise.
In early February we got our job offers, turned in our resignations to our boss the same morning and two weeks later we part of the Multics team at CISL.
As an extra touch of irony, after Multics was cancelled, and some of us from CISL moved back to Honeywell in Billerica - first to build Opus and then other stuff, after a reorganization my boss from the first compiler project and my boss from the System Planning group worked for me.
10 Mar 2016