https://bugs.winehq.org/show_bug.cgi?id=15350
--- Comment #39 from Dmitry Timoshkov dmitry@baikal.ru --- (In reply to Aric Stewart from comment #38)
(In reply to Dmitry Timoshkov from comment #37)
Looking at the japanese keyboard layouts in winex11.drv it seems that the fix should be there instead of dinput. Aric, could you please have a look, or provide more details about the problem you were trying to fix originally?
Hi, I think the earlier comments are pretty clear as to what I was trying to fix. There was a Japanese application using dinput that was depending on the DIK_ values for keys that where mapped differently on the Japanese keyboard than the US keyboard, resulting in unexpected behavior when keys where pressed.
If you mean comments 7 and 13 where you were asking about the test results running your test application this doesn't provide that much information about the problem you tried to fix. On the other hand a comment in the commit message that has caused the regression mentions '[' key, and according to the results I get after loading a japanese keyboard layout in the modified test application, pressing '[' generates same DIK_ code 0x1a for us english, japanese and french keyboard layouts.
I'd suggest to revert the commit that has caused this regression, and open a separate bug report with more details about the problem with input using japanese keyboard layout, and instructions how to reproduce it.