Plan 9 from Bell Labs (opens in new tab)

Shoot, much of the Go lang team came from Plan 9. And they brought a lot of code and ideas with them.

Currently on my mind os `os.FileMode`; here's a comment I wrote in some Go code that interfaces with Python:

    // A StatMode represents a file's mode and permission bits, as represented in Python
    // (i.e. `os.stat()`'s `st_mode` member).  Similar to how Go's `os.FileMode` assigns bits to have
    // the same definition on all systems for portability, Python's `stat` assigns bits to have the same
    // definition on all systems for portability.  And it just so happens that Go's bits match those of
    // Plan 9, and Python's bits match those of the Linux kernel.
    type StatMode uint16

PS: If you weren't aware, Renée French's connection to Plan 9 and Go is largely that she's married to Rob Pike.

The image and image/draw package design[1] is also inspired by Plan 9.

[1] https://blog.golang.org/image-draw

And in roughly the same role (Host FS parts exposed to the VMs) in ChromeOS's Crostini container runtime.

https://chromium.googlesource.com/chromiumos/docs/+/HEAD/con...

In general, 9P2000 is simple and fairly elegant. If you don't need POSIX semantics (and fs metadata), it's really quite simple to implement. For a bit more cost in terms of complexity you can go for 9P2000.u which makes things POSIXy enough if you've got a single client (as you do with VM -> host). This is great if you want to provide virtual filesystems, either proxied from the host or backed by something more exotic. In addition to QEMU, the gVisor folks (https://gvisor.dev/) have made heavy use of 9P (with enhancements) for similar reasons (in addition to Google being generally populated with Plan 9 aficionados).

Then, there's 9P2000.L... This has that "let's just take the filesystem vtable and export it as a protocol" flavor. This works well enough on Linux if you're just looking to translate the calls over into syscalls. I wouldn't recommend implementing a bespoke server for it, though. You're better off just implementing things with FUSE.

Only because I was just reading the Wikipedia page before I posted this :)

  It looks like Plan 9 failed simply because it fell short of being a compelling enough improvement on Unix to displace its ancestor. Compared to Plan 9, Unix creaks and clanks and has obvious rust spots, but it gets the job done well enough to hold its position. There is a lesson here for ambitious system architects: the most dangerous enemy of a better solution is an existing codebase that is just good enough.
  — Eric S. Raymond