Quick Linux insight: A file isn't always bytes on disk. We think of...

A file isn't always bytes on disk.
We think of files as data on a disk, but in Linux, they're an API. A filesystem just needs to handle requests like read and write for a specific path.
This means a filesystem doesn't have to be backed by storage at all.
For example, you could have a filesystem where reading 'random.txt' gives you a new random number every time, with the content generated on the fly. With FUSE (Filesystem in Userspace), you can build exactly that very easily without modifying the kernel, even in a language like Python.
It's a powerful way to make almost any data source accessible to standard command-line tools, or any other tools that expect files.
Check this link for the full text: popovicu.com/posts/how-file…
