Comment by shockeychap
4 years ago
Maybe it's just me, but it always seemed like prohibiting spaces and other special characters was a reasonable way to avoid unnecessary complexity (and the bugs that accompany it) when parsing and navigating directory trees and files.
I'm old enough to remember working with 8.3 filenames in DOS, and while the length limitation was maddening, the space part never was. Then Windows 95 came out and all restrictions were thrown out.
Why couldn't we just have a file system that robustly supports long filenames, including variable length extensions, while prohibiting certain special characters - namely spaces, slashes or any directory denoting characters in files, and characters that have special meaning in regex context? (brackets, asterisk, etc.)
Related: David Wheeler's Fixing Unix/Linux/POSIX Filenames
https://dwheeler.com/essays/fixing-unix-linux-filenames.html
By coincidence, I found another reason just two days ago. A web app lists uploaded files’ names, and (in a rarely used context) lets the user search for them. One user has copied a file name from the web page, and pasted it into the search box, but got no results. Turned out that the file name contained two consecutive spaces, which the browser turns into a single space, hence no match. Every layer between the user and file system can do something unexpected.