Hi folks! It's been a while since my last post in this group, and it feels good to be back with some new FoC-related thoughts: I've been thinking about the tree structure of file systems recently and it turns out that they're limiting and require workarounds for relatively common use cases. Files contain data, but don't have children. Folders have children, but can't store data themselves. What if a file system had "nodes" that could store data AND have children? I've written a blog post about this and would like to hear your thoughts. As I'm not a native speaker in English and not really talented in writing, I'd be interested in feedback regarding the content as well as the general writing style etc.. Thanks in advance and looking forward to interesting discussions! https://fkohlgrueber.github.io/blog/tree-structure-of-file-systems/

index.html

— not to be facetious, and I see you write about it in the article. Of all filesystem woes, data stored in a directory proper is not one that comes to mind. One thing that does come to mind is treating a directory as thought it were a file. I mean macOS packages. https://en.wikipedia.org/wiki/Package_(macOS)

I didn't know about macOS packages, thanks for the hint!

HTML itself is another thing that has this "data at every level" structure, in the form of attributes. It's everywhere. FWIW I think the more interesting problem with filesystem trees is that they're rigidly tree shaped, when tags or 2d tables might sometimes be more useful. :)

@Andrew F yes, file systems being limited to tree structures is an issue. Often, something more general (e.g. a graph) would make more sense. That's something for another blog post some day 😉

If only FSs were honest to goodness trees. Hard links make for big headaches.