On Fri, Sep 27, 2019 at 5:33 PM Eli Zaretskii wrote: > I thought that older versions might perhaps assign different mappings > to the same keysyms. If they don't assign any mappings, then I agree > with you. XKB lists quite a few keysyms as deprecated, but no keysyms we have discussed here are affected. I believe this indicates that they have never been used for another mapping previously, but I haven't checked extensively. However, it is also stated in xkbcommon-keysyms.h that: > * Where the correspondence is either not one-to-one or semantically > * unclear, the Unicode position and name are enclosed in > * parentheses. Such legacy keysyms should be considered deprecated > * and are not recommended for use in future keyboard mappings. This affects four of the keysyms mentioned: > #define XKB_KEY_signifblank 0x0aac /*(U+2423 OPEN BOX)*/ > #define XKB_KEY_leftanglebracket 0x0abc /*(U+27E8 MATHEMATICAL LEFT ANGLE BRACKET)*/ > #define XKB_KEY_decimalpoint 0x0abd /*(U+002E FULL STOP)*/ > #define XKB_KEY_rightanglebracket 0x0abe /*(U+27E9 MATHEMATICAL RIGHT ANGLE BRACKET)*/ For two of these keysyms, I managed to find at least one application that agrees with the current Emacs mapping that we now consider changing: > {0xabc, 0x2329}, > {0xabe, 0x232a}, See https://fossies.org/dox/putty-src/xkeysym_8c_source.html I believe that we should consider carefully whether changing these two mappings could introduce a regression for some use case, e.g. PuTTY/ssh/emacs. My proposed changes are attached.