On Mon, Mar 03, 2025 at 05:36:31PM +0300, Ivan Sorokin wrote: > > Hi Thomas, > > Thank you for your feedback on the proposed patch. I understand your > concerns, and I agree that configurability and avoiding assumptions about > Mod1Mask are important. > > Regarding configurability, I think extending the modifyOtherKeys option would > be the most logical approach. Perhaps a new mode (e.g., modifyOtherKeys=3) > could be introduced. This new mode would behave like mode 2, but also send > the Unicode character (as an escape sequence followed by the character) when > Alt is combined with a non-Latin key and a specific escape sequence cannot be > generated. This would maintain backward compatibility while providing the > desired functionality. > > Alternatively, a separate resource like altSendsUnicode could be considered, > but I believe integrating it with modifyOtherKeys would be cleaner. The main > advantage of this behaviour - that Far Manager (and probably, another console > software) users that uses non-Latin input could bind hotkeys the same manner > as a latin input users. > > What variant do you prefer? integrating with modifyOtherKeys is preferred I'm working on a different change to that feature, but am thinkg about how I'll address this one. > You're absolutely right about Mod1Mask. I was too focused on my own setup. > A more robust solution is needed. What do you think about using XkbGetState? xterm's already getting the modifiers in the X event - I don't see that adding XkbGetState would help. -- Thomas E. Dickey <dickey@invisible-island.net> https://invisible-island.net
Attachment:
signature.asc
Description: PGP signature