Ah neat, this is fascinating. I had never really thought about where they actually came from. I suppose I just assumed it was Java. For anyone else who is curious - the roots of bytecode are in the BCPL compiler and the intermediate O-codes it generated:
It should be noted though that bytecode from that era would often be represented as text rather than binary in storage. If I remember correctly, UCSD p-code was literally just numbers spelled out in text with spaces between them - something that could be conveniently read by just about any language and system around, usually via a primitive like Pascal's own Read function, or BASIC's INPUT.
It makes it that much easier to move from platform to platform when there's no agreement on things that we take for granted these days, like number of bits in a byte etc. But that also makes it harder to recognize what you see when you're used to binary bytecode like JVM or CIL.
