First, the MX700. I bought this mouse in an effort to return to mousing (I had become a trackball user-- a thumb trackball user at that). Almost immediately I found the MX700 to be almost unusable due to mouse jitter and tracking problems. I had tried a few mouse pads and I even tried making my own noise patterns as a little experiment, and those noise patterns and a little almost-precise measurement of jitter-vs-position showed me that it is simply a fact of life that the MX700 set to reasonable sensitivity will pick up certain patterns and oscillate between the edges it finds there, thus jittering.

Now, it is possible that a driver parameter settings change might help. Damping the mouse response so that pixel motion of more than one pixel at the mouse is required to move the cursor, but I resisted this route since I wanted a mouse that was quick and sharp to scoot my cursor across my UXGA desktop. The MX700 had promised this, and I felt let down. I returned to using the trackball, a new Logitech TrackMan Wheel (call me dumb for returning to Logitech, if you like).

Thus, I never actually solved the MX700 jitter problem, unless you count throwing money at it to replace it with a superior product a solution (which I do, for my purposes). If anyone does have any suggestions, I'll take them. For now, I use the MX700-- jitters and all-- when the MX1000 needs to be recharged.

That brings us to the MX1000. Since I don't really require all of the buttons to work for any particular purpose, I'm fine with the mouse properly meshing with XFree86 and the normal buttons behaving, so for now that the mouse is working as expected and is playing nice with the MX700 and my Wacom Graphire 2 all on the usb is sufficient.

I found two pages that pointed me in the right direction. This one and this one were the ah ha moment inducers that I needed to put me on the right track. Until then, I kept thinking that I had to configure some IMPS/2++ or somee such thing to get the extra buttons working in the right order, but evdev support in XFree86 was all that was required.

Either of those links will be useful to you if you're looking for answers to the MX1000 (and other mouse) buttons question, and they have some XF86Config-4 blocks that will definitely put you in the ballpark. Note: Debian Unstable includes evdev patches in the XFree86 build. I can't speak to any other of the Debian distribution versions or any other dist, but my Xfree86 install was waiting for me to enable evdev the whole time. Since I don't swap mice out too often, I'm using the static USB physical port to describe the mouse location for both the MX700 and the MX1000, but this is the next thing I want to play with. Oh, and my Wacom tablet even appears to work with GIMP 2.2 (it even switched between stylus, eraser, and core pointer, which was not working out previously).

When I tested my MX1000 with xinput test, I did get a unique button id for each button press. I think two of the buttons might still have been generating double presses (something like button 4 down, button 10 down, button 10 up, button 4 up, but I'll have to get back to you on the actual button numbers), but for now the mouse simply works, it works well, it doesn't jitter, and I'm happy.

May you find similar levels of harmony in your MX700/MX1000 pursuits...