Regression in lstrcmpiA (occurred in late June, NLS related)

[Dmitry Timoshkov]

"Troy Rollo" <wine at troy.rollo.name> wrote:

> When lstrcmpiA was moved from ole2nls.c to locale.c, (around 28th June) the 
> results of comparisons in some cases became reversed. For example, the 
> underscore now returns as greater than alphabetic characters, whereas it used 
> to return as less than alphabetic characters.

Yes, I'm aware of the problem. Current CX Office CVS has fixes for
all the differences we have found so far. Unfortunately a proper
fix requires a change of the unicode sort weight tables generated
automatically from the unicode.org data base and we (Alexandre, me,
other people at Codeweavers) don't know yet how to make it fit with
future imports from unicode.org.

Unicode weight tables from MS and unicode.org have huge amount of
differences in many absolutely unexpected places...

> The older behaviour was 
> consistent with Win2k.

... and only with Latin1 locale, failing with others.

-- 
Dmitry.