Post #3788835
2026-07-13 15:38 UTC
@david_chisnall@infosec.exchange It might help to understand where I'm coming from if you know that at my job (which is an almost exclusively Python shop and we deal with a lot of historical scientific data that was archived big-endian) it is regularly a problem for me that NumPy exposes the endianness of data on disk to the programmer *without* also giving them the necessary tools to handle it efficiently. You wind up copying entire multi-GB data blocks just to byteswap them.
Replies (1)
-
@zwol@masto.hackers.town 2026-07-13 15:44
@david_chisnall@infosec.exchange Because of this, my thought is we bake into machine language the principle that registers *have no byte order*, byte order is a thing that happens when data moves in or out of memory, and you have to make an explicit choice every single time you do that. And then that gives us the leverage to insist on the same principle at all higher levels. It's a rebuild-the-world idea but that's where all these polls are anyway.