http://bugs.winehq.org/show_bug.cgi?id=5469
------- Additional Comments From juan_lang@yahoo.com 2006-10-07 23:43 ------- Created an attachment (id=2864) --> (http://bugs.winehq.org/attachment.cgi?id=2864&action=view) Tiebreak patch
I think you're right that as long as you stay within the first 256 points, there won't be any chance of different sequences comparing identically.
I think your understanding of the spec appears correct, too, but I'm a bit out of my ken as well. If I understand it correctly, a tiebreak along the lines of the attached patch should do it.
However, the VB runtime passes along flags to VarBstrCmp like NORM_IGNORECASE, and I'm unsure how to deal with the tiebreak when things like that are passed along. See: http://www.unicode.org/reports/tr10/#Multi_Level_Comparison Ignoring the flags altogether is also not correct, that causes some tests to fail.
Any suggestions?