Prev: Fill Form And Get $5000 To Your Home.
Next: How to get USB HID Interrupt ?
From: USH on 27 Jun 2010 21:32

the purpouse was for gaming but not only..
the behavior desired was a linear acceleration under and over any threshold..

games get the pointing signal form i dontknowhow.. bypassing the windows
mouse control panel processing like acceleration..

i heard about a mouse hook..

anyway i use a microsft trackball explorer...
when will it be back to production again!!
im starting to be afraid for some crash of my one..
byby!!



"Ray Trent" wrote:

> On 4/19/2010 10:21 AM, Doron Holan [MSFT] wrote:
> > you really think you want to do this, write a mouse upper filter driver
> > and alter the raw hardware information. remember that there is no
> > floating point math libraries available to you in the kenrel.
>
> Unfortunately, this advice isn't entirely possible to actually implement. The function Windows uses
> is non-reversible, optimized for mice only, and has some really quirky and subtle bugs/"features"
> that I've tried to get fixed in the past with little success.
>
> Perhaps the most obvious of the difficulties is that you can only send integer amounts of motion in
> X and Y, and that quantization is amplified by the Windows transfer functions in non-reversible
> ways. You also can't just send a whole bunch of single mickey movements because of some bugs in the
> way Windows tries to compensate for mouse report rates (I suspect, though of course that's very hard
> to prove without looking at the source), and due to at least one bug I suspect to be a simple
> rounding or overflow error involving small motions specifically in the up & left directions.
>
> The only way to *really* "do what you want" (assuming that's actually a desirable goal here) is to
> have your filter suppress *all* reports, send its packets up to a usermode service that can query
> the position of the pointer, and update it itself to the absolute position on the screen where it
> calculates it wants it. Even that doesn't work perfectly, because there's no atomic way to
> relatively move the cursor other than through the Windows ballistics. Also, you can't use the
> absolute mode interface in the kernel because people can hork the pointer around the screen under
> software control and there's no way to query the position from the kernel.
>
> I've always found it easier to just live with the Windows mouse ballistics, even though there are
> several cases where I feel I could improve on them when using non-mouse devices (like pointing
> sticks and touchpads) having different usage models and constraints.
> --
> Ray
> .
>
From: Maxim S. Shatskih on 28 Jun 2010 13:29
> games get the pointing signal form i dontknowhow.. bypassing the windows
> mouse control panel processing like acceleration..

DirectInput?

--
Maxim S. Shatskih
Windows DDK MVP
maxim(a)storagecraft.com
http://www.storagecraft.com

From: Doron Holan [MSFT] on 28 Jun 2010 13:51
directinput uses HID as well

"Maxim S. Shatskih" wrote in message
news:uJ8cnduFLHA.5472(a)TK2MSFTNGP04.phx.gbl...

> games get the pointing signal form i dontknowhow.. bypassing the windows
> mouse control panel processing like acceleration..

DirectInput?

--
Maxim S. Shatskih
Windows DDK MVP
maxim(a)storagecraft.com
http://www.storagecraft.com

 | 
Pages: 1
Prev: Fill Form And Get $5000 To Your Home.
Next: How to get USB HID Interrupt ?