OSDev.org

SpyderTL wrote:Can you guys think of any technical reason why this would not work?

There's no reason why it wouldn't work. It'd probably be simple getting a prototype going. Getting something that works efficiently with a large data set would be another challenge.

It seems to me that the average user knows little, and cares less, about objects or methods. What they want to know is where they can find that letter they typed up last night, and that is what the OS should serve up. Your operating system sounds like one which might interest computer science aficionados, but nobody else.

sortie wrote:You just reinvented Resource Forks. The concept is interesting and can be useful, but have one main problem: Interaction with dumb tools. What should a tool like cp(1) do when it encounters such a file? What about mv(1)?

One concept I've thought up as an alternative to resource forks is allowing the same filesystem object to be both a file and a directory.

So you might have an executable file /foo/bar/baz, which would also be a directory /foo/bar/baz/. The file itself might just contain executable header information, and the actual code and data segments of the executable might be contained in sub-files (say /foo/bar/baz/code and /foo/bar/baz/data).

This would probably have its own problems with dumb tools, but I find it a rather intriguing concept.

mathematician wrote:It seems to me that the average user knows little, and cares less, about objects or methods. What they want to know is where they can find that letter they typed up last night, and that is what the OS should serve up. Your operating system sounds like one which might interest computer science aficionados, but nobody else.

Average users don't see the file system data structures at all. What you are talking about is the User Interface, which allows them to quickly find specific data based on certain criteria.

I am working on a file system (minus the files, of course) that will allow the User Interface to provide this functionality to search for and find specific information (Documents, Videos, Images, etc.) quickly by building in the concept of objects, classes and indexes at the file system level. Most operating systems rely on the application to handle all of this data-specific logic. I am trying to move that logic to the OS / File System level, so that it is available to all applications, including the command line shell.