Created attachment 740943 [details] emerge --info 0. On a en_US.UTF-8 locale, 1. Emerge Gvim with Athena GUI (-motif -gtk -gtk2 -neXt), 2. Start a clean instance (gvim -u NONE). Expected: Toolbar labels are displayed properly. Actual: Toolbar labels are garbled. 3. Enter insert mode and try to type an accented letter, say á. Expected: á is inserted. Actual: à is inserted. Interestingly, if the file already contains accented characters, say echo áéíü > accents gvim -u NONE accents The characters are correctly displayed. They only can't be inserted. With "LC_ALL=C gvim -u NONE accents" the text is displayed wrongly (as expected) as áéÃü But the toolbar labels are correct. Also one can always paste accented characters from clipboard (with "+p) just fine. The issue is **not** present when Gvim is built from sources from upstream. In all cases mentioned above I get Warning: Missing charsets in String to FontSet conversion warnings, but again, if built from upstream there is no real problem despite the warnings.
Created attachment 740946 [details] locale
Created attachment 740949 [details] USE flags
See: https://forums.gentoo.org/viewtopic-t-1142874.html