Hi, On Sat, Apr 25, 2009 at 9:58 PM, björn <[email protected]> wrote: > It seems to be incomplete somehow though. There is a new function > called im_set_control() which is never called from anywhere so the > imControl boolean inside the text view helper is always false.
Sorry, I forgot to include patch for option.c. > Moreover, what is the thought behind this? It seems strange to send a > message to set the im state every time you press a key -- in > particular the way it is now the sendImState could be called multiple > times for every key press since keyDown may itself call > doCommandBySelector/insertText/performKeyEquivalent (and even vice > versa). It would be better to send a message with the new IM state > when it changes only (I don't know if the Carbon API allows this > though...I do know that there is a 10.5 only API which does but we > need 10.4 compatibility). > Please look over these issues and submit a new patch when you have one. Ok, I see. I have fixed patches. It will send message from keyDown and performKeyEquivalent when IM state is changed. What about it? Thanks, sakamoto --~--~---------~--~----~------------~-------~--~----~ You received this message from the "vim_mac" maillist. For more information, visit http://www.vim.org/maillist.php -~----------~----~----~----~------~----~------~--~---
0001-Fixed-im_get_status.patch
Description: Binary data
