[H-GEN] A few quick questions

[Andrae Muys]

[ Humbug *General* list - semi-serious discussions about Humbug and     ]
[ Unix-related topics. Posts from non-subscribed addresses will vanish. ]

Russell Stuart wrote:
> [ Humbug *General* list - semi-serious discussions about Humbug and     ]
> [ Unix-related topics. Posts from non-subscribed addresses will vanish. ]
> 
> On Mon, 2003-06-02 at 14:19, Russell Stuart wrote:
> 
>>One thing the manual entry doesn't say is that if you set the threshold
>>to 0, whether the mouse is accelerated or not depends only on how fast
>>you move the mouse, not the distance.  Move it slowly and it is not
>>accelerated.  Move it fast and it is accelerated.  I prefer this
>>"dynamic acceleration" to the traditional behaviour.
>>
>>I remember seeing mentioned in a mailing list post by one of the X
>>developers.  I have never seen it documented anywhere.
> 
> 
> Depending on what people are running, I may of mislead you here.  If so
> sorry.  What I described definitely works on my Red Hat 8.0 box, but it
> may not work for you.
> 
> After posting I went looking for the original source I mentioned.  It
> drove me nuts when I could not find it.  Eventually I found something
> better, which describes exactly what is going on:
> 
>   http://www.spinics.net/lists/xf-xpert/msg06650.html
> 
> In brief, when you set the threshold to 0 with xset the mouse
> acceleration algorithm changes completely - it becomes a polynomial
> based on the raw mouse speed - not linear like the old one.  In other
> words it no longer resembles what is described in the xset man page. 
> According to the page above the new algorithm is based on research done
> at Apple, and from my own personal experience it is _much_ better than
> the old one.
> 
> The kicker is that the post above is dated 2002-Nov-02.  You would have
> to running a modern version of XFree86 for it to be available.  I am
> running a version of XFree86 compiled on 2002-Nov-08, and I doubt the
> proposal would if made it into the official source tree by so I am
> guessing Red Hat applied the proposed patches manually.  Perhaps it made
> it into the standard XFree86 4.3 source.
> 
> For what it is worth, in the source I have the new algorithm calculates
> the mouse movement as:
>     Assume:
>          xset mouse A 0
>          Mouse has moved D pixels
>     Then the calculated movement is:
>          D' = D * D^(A-1) / 2
>     Where ^ means "raise to the power of".
> Its a bit more complex in the code, as D is an (x,y) vector.
> 

Are you certain?  This must be rather impressive code as I am personally 
unaware of any way to raise a vector to an arbitary exponent?

Andrae

-- 
Andrae Muys                       But can it generate *quantum* Haiku
<andrae.muys at braintree.com.au>    error messages, in Latin, where each
Engineer                          line of the error message is a
Braintree Communications          palindrome? -- Mike Vanier on perl



--
* This is list (humbug) general handled by majordomo at lists.humbug.org.au .
* Postings to this list are only accepted from subscribed addresses of
* lists 'general' or 'general-post'.  See http://www.humbug.org.au/