> "Remember how people writing assembly thought compiled languages would rot your brain!"
No, I don't remember that and I've been around awhile. (I'm sure one could find a handful of examples of people saying that but one can find examples of people saying sincerely that the earth is flat.) It was generally understood that the code emitted by early, simple compilers on early CISC processors wasn't nearly as good as hand-tuned assembly code but that the trade-off could be worthwhile. Eventually, compilers did get good enough to reduce the cases where hand-tuned assembly could make a difference to essentially nothing but this was identified through benchmarking by the people who used assembly the most themselves.
If you want to sell us on change, please stop lying right to our faces.
Note also that it took, more or less, a hardware revolution in the form of RISC, to make compilers able to compete. A big piece of the RISC philosophy was to make it easier for compiler writers.
They eventually got there, (and I expect AI will eventually get there too), but it took a lot of evolution.
Starting with the 386, the ISA got a lot more compiler friendly. Up to 286, each register had a specialised task (AX,CX,DX,BX means Accumulator, Count, Data,Base register). Instructions worked with specific regs only (xlat, loop). When 386 and 32 bits happened, the instructions became more generic and easily combinable with any register. I remember people raving over the power of the SIB byte or the possibility to multiply any pair of register. While not RISC, it got clearly more easy for compilers to work with the ISA, and I remember reading in magazines that this was an explicit design intention.
Lots of x86 assembly out there from that time period. Beating the compiler in the eighties and nineties was a bit of a hobby and lots of people could do it.
Modern ISA designers (including those evolving the x86_64 ISA) absolutely take into account just how easy it is for a compiler to target their new instructions. x86 in modern times has a lot of RISC influence once you get past instruction decode.
No, I don't remember that and I've been around awhile. (I'm sure one could find a handful of examples of people saying that but one can find examples of people saying sincerely that the earth is flat.) It was generally understood that the code emitted by early, simple compilers on early CISC processors wasn't nearly as good as hand-tuned assembly code but that the trade-off could be worthwhile. Eventually, compilers did get good enough to reduce the cases where hand-tuned assembly could make a difference to essentially nothing but this was identified through benchmarking by the people who used assembly the most themselves.
If you want to sell us on change, please stop lying right to our faces.