ProQuake + Win7 + maccel

[Magnus]

Whats the fix to disable mouse accel in Win7?

My command line:

glpro436.exe -nojoy -nocdaudio -noipx -mpath %1 %2 %3 -width 2560 -height 1600 -bpp 32 -dinput -exec magnus.cfg

Here's what I run in my config:

m_pitch "0.022"
m_yaw "0.022"
sensitivity "5"
lookstrafe "0"
lookspring "0"
cl_backspeed "400"
cl_forwardspeed "400"
cl_sidespeed "1000"
bgmvolume "1"
volume "0.7"
viewsize "100"
crosshair "1"

scr_conspeed 32700

m_filter 1
sv_aim 2
v_kicktime 0
cl_bob 0
cl_rollangle 0
r_waterwarp 0

r_novis 1
r_wateralpha 0.6
r_mirroralpha 0.3

vid_vsync 0
pq_drawfps 1
pq_maxfps 200

[Patton]

try adding -noforcemaccel -noforcemparms to command line

start/control panel/hardware and sound/mouse/pointer options/un-check pointer precision

[Magnus]

No dice on the command line options.

I'm not messing with windows settings just to play quake, lol.

[MH]

Or use -dinput and you'll get no acceleration too.

[Magnus]

That's already in my command line. It worked fine in XP, not in Windows 7.

[=peg=]

try turning "enhanced pointer precision" on/off in mouse-driver..

[Magnus]

Yea, I'm not messing with anything in windows just to play quake1 with direct input mouse.

If I can't fix this through quake itself, my q1 days are over, haha.

[MH]

I figure what's happening then is that DirectInput is failing to start up. I'm pretty sure that ProQuake still uses the old DirectInput libs from the original source release, whereas Windows 7 has removed support for DirectX interfaces prior to 8. That doesn't mean that DirectInput will fail on all Windows 7 machines by the way; it seems quite hit and miss.

There is another way of disabling acceleration without needing DI, but it needs an engine code change. I know that Baker is aware of it and I know that he has talked about finding a way to integrate it into the code and make it switchable at runtime.

Wait for the next version and nag him so that it goes higher on his priority list I reckon.

[Patton]

I am pretty sure that is turned off by default in "windows settings" anyway. It's not like you are h4xin the registry. Anyways gg magnus.

[=peg=]

Mind you, DirectQ works perfectly fine on win 7 in terms or mouse sensitivity (for me at least).

[Baker]

I wish I had a Windows 7 machine to test out all this mouse point chaos people say they have with Win 7 first hand.

whereas Windows 7 has removed support for DirectX interfaces prior to 8.

Uggg ... could you explain this further? Just the basics, I can usually figure out the rest on limited information.

I thought using DirectInput is mostly initiating the .dll etc, etc.

Has R00k addressed this? I suppose I could look in DirectQ ...

[R00k]

I am using DIRECTINPUT_HEADER_VERSION 0x0800 currently in windows 7,
though i have #define DIRECTINPUT_VERSION 0x0700 (fallback) defined in in_win.c with qrack. I'm having issues with the mouse in 8 whereas the mouse gets stuck in the corner if alt-tabbed and some other issues unrelated to this thread...

Between 1.90 and 2.01 I was having issues with acceleration. Mainly if I alt-tabbed the enhanced pointer would magically turn ON. I verified this by going to the mouse driver settings and the check mark would be checked ON. I'd deselect it, apply settings return to quake, all good. Then alt-tab back out, check the mouse settings in the control panel and poof enhanced precision was back on!

I think the problem i was having was during the re-init of dinput when coming back

		case WM_ACTIVATE:
			fActive = LOWORD(wParam);
			fMinimized = (BOOL) HIWORD(wParam);
			AppActivate(!(fActive == WA_INACTIVE), fMinimized);
		// fix the leftover Alt from any Alt-Tab or the like that switched us away
			Key_ClearAllStates();

			//R00k re-init the mouse after alt-tabbing... (v1.9)
			if ((!mouseactive)&&(fActive))//fixed 2.01
			{
				IN_Shutdown ();
				IN_StartupMouse();
				IN_ActivateMouse ();
				IN_HideMouse ();
			}
			break;

IN_StartupMouse would call IN_InitDinput, which would reload the dinput.dll, which somehow flipped my enhanced pointer back on. Ugh!

Adding this code

	if ((COM_CheckParm ("-dinput")) || (m_dinput.value))
	{
		if (!dinput)//2.01: multiple configs hit this
			dinput = IN_InitDInput();

i think may have fixed that since the dll was still active it didnt need to be reloaded. I was actually adding thiat if (!dinput) because config.cfg would hit that, then autoexec would hit it, then r00k.cfg was setting that too and i was seeing "direct input activated" multiple times on the console. But either way I have no issues with enhanced pointer tripping on anymore.

Edit: actually

			//R00k re-init the mouse after alt-tabbing... (v1.9)
			if ((!mouseactive)&&(fActive))//fixed 2.01
			{
				//IN_Shutdown ();R00k: (8/4/10 build 3083) removing this for now as it's probably not required.
				//IN_StartupMouse();R00k: (8/4/10 build 3083) removing this for now as it's probably not required.
				IN_ActivateMouse ();
				IN_HideMouse ();
			}

Doesnt make a difference. So the commented out lines may not be required. Either way dx8 may be the solution anyways.

[MH]

DirectQ is a little less conservative with the mouse, uses DirectInput 8, selects it as preferably on by default (the technology is so mature these days that it's actually obsolete now, so no need to be cautious about it I reckon) and even if it fails to start up it disables accelerations with Windows messaging too.

[Baker]

R00k,

The Enhanced Pointer Precision thing gets turned on by doing a save of the mouse settings without doing a restore of the old settings first (reading the settings Quake already set and perceiving them as the "true" system settings ... so later it would save the "true" settings not recognizing they weren't the true settings but the ones that Quake had set.).

I also fixed this a few versions ago in ProQuake. It's the restore_spi and get_spi code.

[MH]

The older DirectX (D3D at least) DLLs don't ship with Windows 7. So if you attempted to use Worldcraft 1.5 with it, for example, it would kick and scream about d3drm.dll being missing.