In message <200201242237.g0OMbIr20905@guild.plethora.net>, Peter Seebach writes
:
>Hmm.  So, here's the interesting question, then:  Would it be useful or sane
>to make a psm.c which knows of several variants, and tries them all?  It would
>eliminate "psmi", and we'd have a single driver that would, in theory,
>handle most scroll wheel devices.

It turns out this was much simpler than I thought.

On my laptop, if the BIOS is set to support the internal touch pad, the
external pointing device is not allowed to work like a pmsi.

So, a second question:  Would anyone like to see pms_intelli updated to
support the 5-button mouse protocol?  It's just like the 3-button version,
except that it has two more buttons, and you use a different set of
sample rates.  I think it would be about a 20-minute hack to test, although I
don't have a good way to test it, since I have no mice with over 3 buttons.

Hmm.  Actually, I think I may have one.  So, what do people think?  Would
a psm.c that tests for each of a few types of mice, and tries to pick the
best protocol, be a good thing?

-s