I'll add one more. Around 1982, we had to test printed circuit boards with microprocessors. So we had to come up with a way to control the bus from an external device. Fluke 9020 under GPIB from an HP-1000 computer system fit the bill. However, we did not have a specific compiler and loader for that situation. So we had to write the test program in assembly, and then hand assemble it, load it to RAM, and then run it from RAM. It was fun. If you wanted to run a C program, first you had a compiler translate it to assembly language, then hand assembled it. By the time 1984 came around tools began to appear. Some we wrote, some we bought. To this day I cherish my capability of whenever I look at a piece of code, let it be C, C++(with virtual tables), Swift, Fortran or whatever, translating it into machine language in my head, however faulty it may be, and thus having an insight as to how it operates. One thing HP-1000 CPU had was the indirect bit in memory addresses. That was very interesting.