Although many operating system textbooks offer high-
level descriptions of file systems, few go into sufficient
detail for an implementor, and none go into details about
advanced topics such as journaling.
thx to @amy for rec on BeFS
Although many operating system textbooks offer high-
level descriptions of file systems, few go into sufficient
detail for an implementor, and none go into details about
advanced topics such as journaling.
thx to @amy for rec on BeFS
The directory separator can never appear in a file name because if it did, the rest of the operating system would not be able to parse the file name: there would be no way to tell which part of the file name was a directory component and which part was the actual file name.
IS THAT SO? R U SURE?
dimly beginning to realize that it may be possible no one else in the history of humankind has ever truly parsed a path string
A file system must allow users to
organize their files and arrange them in the way they find most natural.
is a filesystem user-facing? or is it an implementation detail?
The traditional approach is a hierarchical organization.
this is an interface! an engineering argument must first describe the supported access/modification patterns (from the end user, or from other programs the end user invokes), and then explain how the architecture enables these!
cc @dahukanna @yvonnezlam @RuthMalan i'm referring to chapter 2 ("What is a filesystem?", pp 7-20) of this text on the BeOS filesystem and i'm curious if the above rings true to you. i find the attempt at user-centric methodology as well as the performance analysis equally lacking
loading screen tip:
A hierarchy is a familiar concept to most people and adapts readily to the computer world.