there's a bunch of dumb bullshit going on here. I started writing this last night and it's kinda long.
First thing: / as path separator is a Unixism and was not historically standard by any means. It is generally accepted that the first hierarchical filesystem was implemented as part of Multics (an ambitious attempt at a timesharing OS that could serve as a "computing utility" with similar reliability to the phone or electrical systems) and used > as a path separator. DEC's TOPS-20 and VMS both use(d) a . (and have other similarities as well), MIT's homegrown ITS used a ;. Even among "modern" operating systems, this continued; Mac OS prior to OS X used a :. Beyond that, many of these systems had other niceties; TOPS-20 and VMS, for instance, have structured pathnames where the directory is a single delimited component, wrapped in <> and [], respectively. Many systems include some sort of drive or volume specifier; Unix's single tree is extremely unusual, historically. Finally, many operating systems, particularly microcomputer systems like the mini-OS built into the Commodore 1541 disk drive, Digital Research's CP/M and early PC-DOS, but even some early mainframe OSes, had no support for hierarchical filesystems at all, so there was no path to separate.
So how does this relate to Windows? Well, when the IBM Personal Computer was in development, besides the home computer market (C64, Timex Sinclair, BBC Micro, TRS-80, etc) there was already a thriving-ish market for business machines based on the Intel 8080/8085 or Zilog Z80 CPUs, which ran CP/M. IBM had no operating system for the IBM PC at the time and approached both Digital Research about porting CP/M to the PC, and Microsoft, who at the time were known for selling the Microsoft BASIC interpreter that was already licensed to every computer manufacturer that would take MS's calls.
There's a lot of mostly bullshit legendry around how this all went down but the important part is that MS did not have an OS to sell at the time, so instead they acquired a CP/M semi-workalike called 86-DOS from Seattle Computer Products. CP/M didn't have directory paths, but instead had drive letters (named devices like this being a typical way to address different volumes on basically every non-Unix system; CP/M just had a particularly simple-minded approach), so each filename, fully qualified, was something like A:FILENAME.TXT, for a file named FILENAME.TXT on the first floppy drive. 86-DOS copied this and a lot of other CP/M-isms.
The early PC DOS utilities settled on / as a switch specifier for passing parameters to commands (eg, if you want a directory listing and you want the listing to pause after each "page", or screen, of listing you would type "DIR /P"). There's a lot of confusion about where this came from. It didn't come from CP/M or IBM's mainframe OSes, none of which used this convention, though that's commonly said to be the case. However, DEC systems (like TOPS-10 and TOPS-20) did, and MS did a lot of early development of eg Altair BASIC on a PDP-10 so it's likely that's where it's from.
So for PC DOS 1.0 we had drive letters pointing to named floppy drives, and forward slashes as command flags and that was fine. DOS 2.0, however, introduced a hierarchical filesystem (a modification of the really simple fs used in 1.0 that was called FAT, and which in a modifier 32-bit form is still commonly used today) so as to be able to support hard drives. The obvious choice for the path separator is a forward slash; it's 1982/1983 now and Unix has unequivocally won, such that DOS 2.0 is consciously taking on a number of other Unixisms like pipes, file descriptors and environment variables. However, we use forward slashes pervasively for command switches. So instead we use backslashes. We don't discard the drive letter system for many reasons (backwards compatibility and the fact that most PCs, even those with hard drives, never had a stable "root" FS because people were constantly booting from floppies that we're going to be swapped around for various reason) just adapt it; assign letters to all floppies first, then to all hard drives. IBM branded hardware only ever supported two internal floppy drives, and two drives was typical on clones, so C: was usually the first hard drive, and so as of DOS 5.0 that was hardwired in which is why your system drive is C:. Windows up to 3.11 was all just a shell on top of DOS (and Win9x was still mostly that) so Windows inherits the same structure too.
Nowadays, modern Windows supports / as a path separator and supports Unix-style volume mounts as well, but drive letters are still a primary designator and mostly storage devices are still addressed this way, but if you have a lot of partitions you can set up mount points for the ones you can't assign. In the old days you just couldn't have more than 26 total drives but there was no hardware that could support that anyway without stupid partition tricks, so you'd only run into the limit if you were doing a lot of network drive mapping or going nuts with SUBST, which meant that in practice it was mostly something systems administrators worried about.
Those long paths starting with \ are called UNC paths and are a sorta-kinda single tree naming system that allows all "devices" to appear in a single tree. "\localhost\c$" is actually referring to your computer "over the network" and connecting to the autogenerated full disk share for the C: drive.


.png)