El 02/02/2006, a las 23:15, Julio M. Merino Vidal escribi=F3:

>
> El 02/02/2006, a las 21:17, Hubert Feyrer escribi=F3:
>
>>
>> On Thu, 2 Feb 2006, Julio M. Merino Vidal wrote:
>>> Please comment.
>>
>> My Toshiba Portege 3440ct has two such "scroll" buttons above the =20
>> "normal" mouse buttons, too, and they just work, including =20
>> autorepeat if I keep them pressed. I guess the hardware is sending =20=

>> single keypresses.
>>
>> I'm not sure I understand your patch correctly - is it for =20
>> hardware that doesn't send repeated keypress events (scroll up/=20
>> down), and doing that in software?
>
> Yes, exactly.
>
> Now that you mention this, I think of a way to simplify the whole
> thing I did in wsconsctl(8).  Maybe, instead of allowing auto-repeat
> on any button, just tell the kernel which are the two buttons to
> be used for scrolling and let it simply DTRT.  E.g:
>
> $ wsconsctl -w -m scroll.up=3D3 scroll.down=3D4
>
> It couldn't be as flexible as what I did, but maybe it is more than
> enough and far simpler for the end user to understand.  (I can't come
> up with why you'd need autorepeat in other buttons, but who knows.)

Hmm... well, I'd also have to take into account four-directional
scrolling, so add scroll.{left,right} to the mix.

But, in fact, I like this idea a lot.  It'd free the user from timing
details and the kernel could do better to simulate the scrolling.
(I'm thinking about progressively accelerating the speed while
helding a button down.)  I could Just Work (TM) :-)

--=20
Julio M. Merino Vidal <jmmv84@gmail.com>
The Julipedia - http://julipedia.blogspot.com/