Thanks Dan. I added my findings - but I don't have permission to change the status to confirmed, and I don't have time to investigate any more now. On Fri, 14 Oct 2005, Dan Kegel wrote: > Filed as http://bugs.winehq.org/show_bug.cgi?id=3579 > -- Bug confirmed. Further information: The text in question is inside an edit control (dlls/user/edit.c). A patch in February added checks inside EDIT_EM_ReplaceSel to make sure that the text will fit inside the available box. Before this, there was no such check. http://www.winehq.org/pipermail/wine-patches/2005-February/015979.html In the current code, this check is ignored if honor_limit is false. If honor_limit is true and the replacement (or new) text appears to be too long, the edit control is left unchanged. The message in IBasic is calculated as taking up too many lines, which is why it does not appear nown, but did appear before the line-number check was added. Details about this particular message: the edit control has its top at pixel 0 and its bottom at pixel 48. The line_height property is 16, The calculated number of lines required is line_count=5, and the calculated number of lines available is vlc=3. Because 3 < 5, EDIT_EM_ReplaceSel concludes that the message is too long and doesn't display it. If I turn off the length checking to make the message appear, it requires only 4 lines, and the actual offset of one line from the previous one is 14 pixels. >From these details, there are three distinct things that look wrong to me. 1. Each line requires 14 pixels, but line_height=16. Why don't they match? 2. The message is calculated to require 5 lines, when it really only needs four. I think this is a problem in EDIT_BuildLineDefs_ML. 3. If #1 and #2 were fixed, we'd need 14 pixels * 4 lines = 56 pixels, and the edit control only has 48 pixels, so the length check would still fail. Maybe something else is wrong? _______________________________________________ wine-users mailing list wine-users@xxxxxxxxxx http://www.winehq.org/mailman/listinfo/wine-users
