[mdetect?] detect number of mouse buttons

[Daniel Borgmann]

Currently Ubuntu only supports up to three mouse buttons and a wheel. To enable
a fourth mouse button, one has to change the mouse protocol from "ImPS/2" to
"ExplorerPS/2" (this works with my Logitech Dual Optical, so it's not MS
Explorer specific) and add the option "Buttons" with a value of 6 to the
XF86Config. I don't know what would have to be done for more complicated mouses
with more buttons and/or wheels.

[Daniel Borgmann (spark-mailinglists)]

It also works with setting "Buttons" to 99 (but not with omitting the option).
Makes me wonder if this could be an option to support an unlimited number of
buttons. Potential problems are:
- I doubt that all mouses are supported by the ExplorerPS/2 protocol. Could this
be autodetected?
- By default the scroll events are mapped to buttons 4 and 5 in the XF86Config.
Basically all HOWTOs on enabling four and more mouse buttons tell, that you have
to remap this if you use more buttons. Interestingly I didn't have to do it this
time with Ubuntu, which makes me wonder if the problem was fixed or if I'm just
lucky. If it's not fixed, then this problem could be very hard to solve without
autodetection of mouse cababilities.

[Matt Zimmerman (mdz)]

Is this a PS/2 mouse or a USB mouse? If PS/2, do you use /dev/psaux or
/dev/input/mice in the config?

[Daniel Borgmann (spark-mailinglists)]

It's a USB mouse and I deleted the double entry from my XF86Config. I just
tested the above settings with a plain three button wheel mouse (Logitech
Optical, USB) and it also worked perfectly.

[Fabio Massimo Di Nitto (fabbione)]

hmmmm I need a reliable way to know which protocol to use. I am not happy to
change default protocol or adding options like Button just because they seem to
work.
There are too many different config around the world to be sure 100%.

We can discuss this problem again for Hoary, and see if we can handle everything
with little tools that will ask max one or two questions and reconfigure
different bits and pieces.

[Fabio Massimo Di Nitto (fabbione)]

Daniel, no more information have been added to this bug. I am still
not convinced that we should change default protocol.

We will review this problem for Hoary release.

[Fabio Massimo Di Nitto (fabbione)]

Matt, please respect the way in which I handle the bugs. LATER means that it
will be fixed LATER.
The status is not documented in bugzilla, but my understanding is that they will
be reopen after warty release
and clearly targeted for the next one if possible. Nobody ensures us that we
will have a tool to fix this problem
for hoary, since it involves a default protocol change.

[Fabio Massimo Di Nitto (fabbione)]

As agreed with Matt on irc.

[Daniel Stone (daniels)]

This is pretty much impossible -- we have to stab and guess. Maybe this is a
job for the all-singing, all-dancing, hardware database ...

[Daniel Stone (daniels)]

This bug has been marked as a duplicate of bug 10665.