moused linear acceleration drops

[Harald Schmalzbauer]

Am Freitag, 15. Juni 2007 schrieb Oliver Fromme:
> Harald Schmalzbauer wrote:
>  > 3 months ago moused was extended by a _very_ nice feature: Dynamic
>  > acceleration. It works really fine.
>
> Nice to hear that (I wrote the patch).  :)

I knew, I read the PR because I was interested who took the pity on this long 
standing disadvanted compared to some othe operating system... So you alredy 
got my thanks in mind :)

>  > But the linear ac(de)celeration seems to 'drop' moving information.
>  > I have a 1600dpi mouse which is far to quick by default.
>  > When I use 'xset m 1/3' the speed is fine and also every smallest
>  > movement gets recognized.
>  > When I use '-a 0.33' with moused (besides -A 1.6/2 with xset m 1) small
>  > (and slow) movement's are not reported. I have to move at some minimal
>  > speed to get the cursor moving.
>  > So it seems that info gets droped.
>
> It's probably a rounding problem.  I'll have a look at it.
> My mice and trackballs aren't that sensitive, so I didn't
> notice such a problem.
>
> I assume that there's no such problem if you don't use the
> new -A option, right?  (i.e. no regression, I hope.)

Hmm, if I don't use -A (dynamic acc) the problem still persists, the linear 
acceleration is droping info, hence the culprit. I don't have a problem with 
dynamic acceleration when the linear is disabled, so it can't be a regression 
by your patch I think.

Thanks,

-Harry