http://bugs.winehq.org/show_bug.cgi?id=6515
------- Additional Comments From michael.teske@swissrisk.com 2006-28-11 02:20 ------- This does not fix the problem. But this might reveal a clue. I added some FIXMEs to output some variables:
fixme:richedit:ME_Repaint ME_Repaint had to call ME_WrapMarkedParagraphs fixme:richedit:RichEditANSIWndProc EM_EXLIMITTEXT editor=0xe32860 wParam=0 lParam=2000000 fixme:richedit:RichEditANSIWndProc WM_NCCREATE: editor = 0xe331c0 fixme:richedit:RichEditANSIWndProc WM_NCCREATE: editor = 0xe33b20 fixme:richedit:ME_Repaint ME_Repaint had to call ME_WrapMarkedParagraphs fixme:richedit:ME_Repaint ME_Repaint had to call ME_WrapMarkedParagraphs fixme:richedit:ME_InsertTextFromCursor ME_InsertTextFromCursor editor=0xe331c0 editor->nTextLimit=32767 fixme:richedit:RichEditANSIWndProc EM_EXLIMITTEXT editor=0xe33b20 wParam=0 lParam=2000000 fixme:richedit:ME_InsertTextFromCursor ME_InsertTextFromCursor editor=0xe331c0 editor->nTextLimit=32767
It seems that some repaint operation changes the editor pointer, so that ME_InsertTextFromCursor and EM_EXLIMITTEXT go on different editors.
Bug 6148 might have the same cause...