Tom Spear wrote:
Hi all, I was looking at bug 3747, about the game Motorhead. This game crashes because IDirectInputAImpl_GetDeviceStatus always returns DI_OK due to being a stub.. The reporter came up with a <10 line hack to fix it, but a proper implementation would be better. There may be other games that suffer problems due to this not being implemented.. Perhaps someone could look into the feasibility of this as a GSoC project, or possibly could come up with a minimal non-stub implementation to keep games like that from crashing just because of a joystick? I made a bug for the stub function.. Bug 7770..
No hack will work there. Or you risking braking all programs/games that use DInput. It should be properly implemented or not at all. IMHO it's too small of the project for SoC. Fixing all of DInput - might be. That would include merging both joystick interfaces together, adding proper configuration options to winecfg. Fixing other remaining mouse, keyboard and joystick related bugs in dinput.
As far as mouse goes - that would require someone like Alexandre himself to fix winex11drv properly. Or completely different approach to this, like using /dev/input/event. However some one mentioned that it doesn't work (not allowed) for primary pointer device.
On the side note, if anyone interested, I might have barely enough time for mentoring.
Vitaliy.