When I was roaming the halls of UC Berkeley 25 years ago, the use of countable "codes" was idiomatic in the numerical computing sphere at the intersection of co

When I was roaming the halls of UC Berkeley 25 years ago, the use of countable "codes" was idiomatic in the numerical computing sphere at the intersection of computer science, mathematics, and parallel computing. A code was more or less a program or application. A frequently linked idiom was "kernel" to refer to the inner-loop of a particular numerical simulation, generalizing a "filter kernel" (matrix of coefficients) in signal processing/convolution. I have heard similar usage from academics working in various US national labs, so it was not confined to a single coffee klatch. Some of the professors and postdocs using it back then are probably lab directors and program managers by now. I can easily imagine that this usage would be widespread among academic and federal lab computing environments. Like many kinds of jargon, it is both more precise in its meaning when used properly, but also what you might consider a "dog whistle" used for virtue signaling.