Oh. Shit, first I didn’t know what a nibble is, now I didn’t know what a word is … I’ll just show myself out and make sure I never try to write any assembler code.
Actually that does sound best… but not because of anything to do with you, that advice just applies to all humans for the last basically twenty years. So it’s nothing to do with AI slop, but rather compilation got SO FREAKING GREAT that in like 99.9% of cases it does better than a human.
Ironically, one of the reasons for that is that concepts like “word” varies between different architectures (like like “byte” too - e.g. you’ve likely heard the distinction between 32-bit vs. 64-bit systems). As the other person said it is usually 4 bytes but other times it can be like 2 (in super old systems I suppose) or more than 4 in like super powerful ones (128-bit ones maybe?).
And compilers keep track of ALL of those minutiae, as well as a bunch more stuff too. Like a small(-ish) loop that only runs a couple to few times is far more efficient to just be duplicated in the assembly code, rather than have the GO-TO procedure jump back to its beginning spot. And also it helps immensely to feed things to the chip based on what the various cache sizes are - L1, L2, L3 (though not all chips have all… or even any of those).
So yeah, it’s best to leave compilation to the computer to handle, thus freeing up the programmer to handle the overall logical flow and deal with other higher-order effects like code reuse and documentation and stuff.
Oh. Shit, first I didn’t know what a nibble is, now I didn’t know what a word is … I’ll just show myself out and make sure I never try to write any assembler code.
Actually that does sound best… but not because of anything to do with you, that advice just applies to all humans for the last basically twenty years. So it’s nothing to do with AI slop, but rather compilation got SO FREAKING GREAT that in like 99.9% of cases it does better than a human.
Ironically, one of the reasons for that is that concepts like “word” varies between different architectures (like like “byte” too - e.g. you’ve likely heard the distinction between 32-bit vs. 64-bit systems). As the other person said it is usually 4 bytes but other times it can be like 2 (in super old systems I suppose) or more than 4 in like super powerful ones (128-bit ones maybe?).
And compilers keep track of ALL of those minutiae, as well as a bunch more stuff too. Like a small(-ish) loop that only runs a couple to few times is far more efficient to just be duplicated in the assembly code, rather than have the GO-TO procedure jump back to its beginning spot. And also it helps immensely to feed things to the chip based on what the various cache sizes are - L1, L2, L3 (though not all chips have all… or even any of those).
So yeah, it’s best to leave compilation to the computer to handle, thus freeing up the programmer to handle the overall logical flow and deal with other higher-order effects like code reuse and documentation and stuff.
Well technically
https://en.wikipedia.org/wiki/Syllable_(computing)