Multiple mice in Windows 2000

[jerryjanis]

I just discovered an open source development project called CPNMouse that makes it possible to use multiple mice in Windows 2000 and XP without using RawInput.  He created a "filter driver" which (I believe) acts like the usual driver, but also lets you use a multiple mouse API that he has written as a supplement for the driver.

Read about it on this site:
http://cpnmouse.sourceforge.net/

Before he started writing code, he wrote a publication entitled "Supporting Multiple Pointing Devices in Microsoft Windows" in which he references Mame Analog+!  Check it out here:

http://klafbang.dk/personlig/publications/mouse.pdf

As of right now, it works with Windows 2000, XP, and 2003.  If a driver can be written for Windows 9x, then this system could potentially work for all the important versions of Windows (95, 98, ME, 2k, XP, 2k3).

The biggest difference between this solution and the current solutions (DirectX and RawInput) is that it requires that a different mouse driver is used than the default Microsoft one.  I've been playing around with it, and it seems really cool so far.  It supports Serial, USB, and PS/2 mice.  It works with lightguns and mice, and I'm able to compile the api demo program using the Mame compiler environment (MinGW/GCC).

I put up a web page where I plan to keep track of my progress while I play around with CPNMouse.  You can check it out here:

http://jstookey.com/arcade/cpnmouse/

I'm curious to know people's thoughts.

[Lilwolf]

Is there any problem for people who have to use a specific mouse driver (does it wrap another driver... or itself is a driver?)

But will it solve much over the current seperate solutions?  Other then getting it to one code base for both?

[jerryjanis]

Is there any problem for people who have to use a specific mouse driver (does it wrap another driver... or itself is a driver?)

I don't know.  I emailed this question to the author.  I've never used specific drivers for any mice.  Have you?  Do you have any idea of how much of an issue it is?

Here's an image taken out of the pdf document describing cpnmouse, if that helps to confuse you any:  

But will it solve much over the current seperate solutions?  Other then getting it to one code base for both?

Ehhh, no not really - just the fact that there could be consistent support for nearly all versions of Windows.  Also, I sort of believe that having choices is a good thing.  If somebody has trouble getting one method to work, having a second method to try could be a good thing.  I know, that's not really the Windows Way, and for good reason - it makes things more complicated.  That's where I'm wondering if this solution is a good thing or not.  I'd love hear opinions on the matter.

[jerryjanis]

I have gotten a response from the author:

> Is there any problem for people who normally have to
> use a specific mouse driver?

There shuold be no problems, as the normal mouse driver is still loaded, CPNMouse just loads above it in the driver stack, and filters out unwanted events.

When a completely new bus for mice appears, a new .inf file will have to be developed, but that should be all.

[Lilwolf]

Cool!  I only have one trackball that doesn't use normal drivers...  But that one isn't currently being used on my arcade machine.  But someone might find a problem with it..

Anyway.  I think the question is would someone keep custom drivers loaded all the time on their system... Instead of downloading the analog+ version specific to their system.  I don't think there are too many which use BOTH xp and 98 and multi-mice on their systems.  So putting development time might not be all that useful.

But then again.... then xp 'fix' isn't supported by microsoft.  So you might find it not working in 2003... and win64.  So this might be a good solution for the long term.