On 2026-10-30 16:08, daniel Azuelos wrote: > How is vi able to display the 2 formats of coding utf-8 correctly > without any specific configuration?
both constitute different byte sequences that vi just hands off to the coresponding components. When you specify opening a file in vi, you're telling it the filename (composed or decomposed) and it hands that off to the filesystem OS calls. When vi sends that (composed or decomposed) to the display, the display renders that sequence of (composed or decomposed) bytes accordingly, which might result in the output *appearing* the same, even if the input was different. > Is there any configuration to make mutt read the 2 kinds of file name > encodings with utf-8 exactly as vi is able to display them? I don't know the internals of mutt well enough to answer with great authority, but it might depend on how it's getting the filenames: are you typing them, or tab-completing them from the disk, or from a config file (that would have the correct composed/decomposed form), or from mutt's file-browser? I would expect that mutt will happily take filenames it receives from the filesystem OS calls (composed or decomposed) and use them, handing them off to terminal display routines that display the composed or decomposed byte-streams (hopefully appropriately). This would likely include tab-completion of filenames or things in the file-browser. But if they're user-supplied filename byte-streams, whether hand-entered at a prompt or hand-entered in a config-file, you'd be rsponsible for ensuring that they match the composed or decomposed byte-stream that forms the name on disk. -tkc --
