On Holy Wars and a Plea for Peace (1980)

Post by kstenerud » Tue, Aug 03, 2021, 3:41 PM UTC

bear8642 wrote: post #11

kstenerud wrote: post #7

I wrote up a blog post about the strengths of each endianness: https://technicalsourcery.net/posts/on-endianness/ Short short version: It's all about where your imaginary "margin" is. Little endian works best with "upward growing" data, where the data grows from the low bits into the empty high bits (for example integers). Big endian works best with "downward growing" data, where the data grows from the high bits to…

Thanks for blog, very useful analogy. With the odd/even detection section, do you know why the bytes aren't also small to big in Little Endian? That is why isn't the one's bit the leftmost bit?

It's just the convention we've adopted for bit positioning. To the programmer, it doesn't matter since you never actually see sub-byte data, so we just conventionally use the same meaning of shift "left" and shift "right" opcodes regardless of byte-endianness or how the CPU itself arranges the bits.

So if we had followed the original Indian endianness, then a shift "left" would divide by 2 and a shift "right" would multiply by 2. All of our troubles stem from not swapping the digits around when we adopted the Indian/Arabic number system in the 10th century - we had to keep up the convention at first because it was compatible, and later because it's "natural".

Re: On Holy Wars and a Plea for Peace (1980)

Re: On Holy Wars and a Plea for Peace (1980)