“Accurate” is probably not the correct word anymore. It was when technical limitations dictated power-of-two capacities. Commodore 64 came out with 64 kiB = 216 B of memory, and FAT32 cannot handle file sizes ≥4 GiB (232 B). However, RAM/ROM/Flash chips manufacturers no longer make exclusively powers-of-two capacities, instead opting for (decinal) GB to save 7 % of the cost (and other fake capacity shenanigans). I prefer binary too but the two unit systems can coexist, people just need to label them correctly.
Recommended units for data have been mibibytes (MiB), gibibytes (GiB), etc. for a few years now
They’re more accurate because they use powers of two (actually 1024 instead of 1000)
“Accurate” is probably not the correct word anymore. It was when technical limitations dictated power-of-two capacities. Commodore 64 came out with 64 kiB = 216 B of memory, and FAT32 cannot handle file sizes ≥4 GiB (232 B). However, RAM/ROM/Flash chips manufacturers no longer make exclusively powers-of-two capacities, instead opting for (decinal) GB to save 7 % of the cost (and other fake capacity shenanigans). I prefer binary too but the two unit systems can coexist, people just need to label them correctly.
Ah I didn’t know that