I feel like a lot of these issues will disappear for a long time, after the switch to 64 bits: bits max value 8 bits 255 16 bits 65_535 32 bits

I feel like a lot of these issues will disappear for a long time, after the switch to 64 bits: bits max value 8 bits 255 16 bits 65_535 32 bits 4_294_967_295 64 bits 18_446_744_073_709_551_615 128 bits 340_282_366_920_938_463_463_374_607_431_768_211_455 And that applies to most things: time stamps: 64 bits can fit hundreds of billions of years (longer than the age of the universe) IP addresses: an address for every device, no more CGNAT, the block assignment sizes are mindboggling (/64) DB primary keys: it's unlikely that most systems out there will need more than 18 quintillion records coordinates: whether you're working on video game worlds, or simulations, this is probably accurate enough for floating point calculations and storage hardware: millions of TB of RAM is also likely to be sufficient when using 64 bit OSes I feel like once the switch is made in most places, we're all going to have a more relaxed experience for a while. For example, currently if I try to create a game where the game engine uses 32 bit floats, then after a certain distance away from the scene root point, the physics break down a bit due to the decreasing accuracy. That's how you get the complexity of floating origin systems where you reset the camera/player back to the root point, alongside everything else relative to them, after they've moved a certain distance away. With 64 bit values, there would be no need for the complexity of such workarounds. And it's pretty much the same for computer networking (the whole NAT/CGNAT thing), dates and everything else, where you actually run into the limitation of 32 bits not being enough.