http://bugs.winehq.org/show_bug.cgi?id=18424
--- Comment #13 from Mr Nobody limited_choice@hotmail.com 2010-12-30 14:33:42 CST --- (In reply to comment #9)
If you can find even a few games where it does work ok, thats at least a starting point to see if there are still any issues.
You've no idea how accurate that turned out to be ;) First up though, confirming my memory didn't fail me when referring to past experience in linux - using a wine-1.3.10 build and the gog.com version of IL-2 Sturmovik 1946, this app simply does not see the USB stick at all. On the same machine, with a PCI emu10k1 soundcard with the older gameport style interface, I have connected a SkyMaster RapidShot analog joystick (4 button, 2 axis + throttle wheel. HAT). This device *must* appear @ /dev/js0 to be seen by IL-2, so you've got to unplug any other joystick HIDs you have (and perhaps rmmod joystick) to claim /dev/js0 -- IL-2 sees this gameport connected joystick no problem at all. Why this is so, I cannot say, but I'd always presumed it was due to snd_emu10k1gp driver being present in the device tree. If they're getting USB joysticks to work with IL-2 in Windows, you'd need question whether or not any 3rd party drivers are being used..ie; in my mind, I doubt it'd be hard to create a software gameport to emulate that device hardware .. 'who knows what goes on in Windows'.
I finally gave up on looking for an open source program or other free flight(sim) genre titles, after the 11,2 spat the dummy about that notion 3 times (couldn't get the apps to run)...so instead, scrub flight, lets go driving instead (Flatout on Steam) - hey presto, there's the Saitek ST-90 as default choice, the Logitech Dpad, 'Controller' (xbox pad) and keyboard, all as they should be. Going to configure it reveals a little trap - Usboverdrive has claimed it, and it sent the app keyboard events...(just as it should really) ..turn off Usboverdrive, and now Flatout sees the 'real' joystick device, all it including the digital-analog throttle (although useless in a driving sim, it would be at home in a flight/other sim)...in fact, it works as good on the Mac as it does in linux ... the support is definitely there, and ostensibly found to be working 'just as it does in Windows' in this test case, again using a build of wine-1.3.10 via macports.
Little doubt I could find other apps equally disposed if I kept on looking and checking...but having established this much, how best to proceed in isolating any anomalies? I believe I may have a test case scenario, based on this little excursion....ie; those apps mentioned previously where in OSX they couldn't see the xbox pad but the same apps -could- see an xpad in linux ; maybe I'll get lucky and we can get a trace wrt simple USB joystick on both working and non working situations and find the gremlins.
Should I be looking at anything more other than dinput in the way of debug channels at this stage?
Kudos to all for the interest and feedback.