Why windows still boots from drive c: in 2024
Three letters have haunted every Windows installer since 1983: C:. While the rest of the alphabet moved on, that single character remains frozen in time, a fossilised shortcut that still dictates where your operating system breathes. Microsoft could have killed the relic decades ago. It didn’t. The reason is part engineering, part superstition—and entirely about the fear of breaking the world.
The ghost of drives a: and b: still roams
Flip open any PC built after 1995 and you’ll notice something odd: the alphabet starts at C. The empty seats—A: and B:—were once reserved for 5¼-inch floppy drives. CP/M, the microscopic OS that preceded MS-DOS, hard-coded those letters in 1974 so that boot code could find the first disk without asking questions. When IBM cloned CP/M to create its Personal Computer, the lettering scheme came over like a virus. The 1981 IBM 5150 shipped with one or two floppies; the 1983 XT added a 10 MB hard disk and simply grabbed the next free letter. Overnight, C: became shorthand for “permanent storage” and the meme calcified.
Microsoft could have normalised the sequence at any point. Windows 95, XP, even 11—each release was a chance to rename the system volume to something rational like S: or Sys:. But every time engineers ran the migration simulator, the same horror story unfolded: 40 years of legacy software, from dusty accounting binaries to niche medical drivers, still harcode paths such as C:WindowsSystem32. Swap the letter and those programs don’t throw polite errors; they implode silently, taking entire production lines with them. The cost of patching outweighs the aesthetic win. So the letter stays.
Disk letters are not labels, they’re contracts
What looks like a drive name is actually an ABI promise. Change it and you break the contract. That’s why the Disk Management snap-in greys out the option to alter the system volume. Sure, registry hacks exist—load offline hives, hex-edit mounteddevices, pray. But the moment you reboot, Windows Update, Defender, even the Xbox overlay start screaming: path not found. Enterprise customers learned this the hard way during the Vista era; a single re-lettered fleet cost one logistics firm €3 million in downtime. After that, Microsoft hardened the block. Today, even the Windows Preinstallation Environment refuses to cooperate if you try.
USB sticks and SD readers happily grab D:, E:, F:, perpetuating the illusion that letters are arbitrary. They’re not. The kernel still enumerates volumes in DOS order first, then NT namespace. Plug in enough removable media and you’ll eventually collide with the ghost range; Windows will politely skip A: and B: even if nothing owns them. It’s a bow to 1983, baked into the Plug-and-Play manager.
Curiously, A: and B: aren’t extinct. VMware will let you mount a virtual floppy to A: for retro gaming. Industrial loom controllers in northern Italy still demand B: for nightly firmware dumps. Microsoft keeps the slots warm, just in case. Meanwhile, the user folder—another immovable object—can’t be renamed either. Try it and OneDrive, Office, Adobe Creative Cloud throw tantrums because %USERPROFILE% expands to a hard path. The same inertia, different coat.
So every new laptop ships with a 500 GB NVMe SSD proudly wearing a badge that should have retired 30 years ago. We laugh, we complain, we forget—until the day some rookie sysadmin decides to “clean up” the alphabet and discovers why the veterans drink. C: isn’t a letter; it’s the weight of legacy pressing down on the present. And legacy, unlike disks, never dies.