Re: Mouse acceleration

[taob@r-node.hub.org (Brian Tao)]

In article <jmk3.734393295@crux1.cit.cornell.edu>, Jay Krell writes...
> 
> I've seen this suggested before. The thing is, you can just multiply all
> mouse delta, 'cause that would make many pixels unaccessable. If someone
> want to write this (me?), you'd have to watch the deltas, if consecutive
> ones are large (mouse is moving fast), increase them even more.

    All you really have to do is watch the deltas and transform the
values according to a lookup table.  If the delta is less then 3 pixels,
move the mouse at face value.  Between 4 and 5, add 1 to the delta;
between 6 and 7, add 2 to the delta; etc.  You can write this as a CDev
and provide a default "acceleration curve", and let the user modify it
to his tastes.  Actually, it isn't *that* big of an issue since the most
you ever have to move is 640 pixels anyway.  Hmmm... come to think of
it, this could come in handy with Jumbodesk...
-- 
Brian Tao:: taob@r-node.hub.org (r-Node BBS, 416-249-5366, FREE!)
::::::::::: 90taobri@wave.scar.utoronto.ca (University of Toronto)
][ Read news.groups for the discussion on csa2 reorganization! ][