On Thu, May 24, 2007 at 01:58:27PM +0300, Paul Sokolovsky wrote:
> Hello Damien,
> 
> Thursday, May 24, 2007, 1:23:39 PM, you wrote:
> 
> > Paul Sokolovsky wrote:
> >>         Secondly, if you looked at the driver, such an event
> >> explicitly generated at the end of touch, see line 197. And that's
> >> arguably correct, as driver must report when the touch was finished,
> >> and it's not always possible to do that with the last reported correct
> >> coordinates. So, it seems like bug in evtouch which tries to interpret
> >> coordinates of lifted pen (??). Of course, it could be that issue is
> >> still in ts-adc-debounce (and all other drivers we have been having,
> >> on which it is based), but please prove that.
> 
> > Any driver using the input "mouse" interface
> 
>         Touchscreens don't use "mouse" interface.
> 
> > must assume that every 
> > coordinate sent by the driver are correct. So ts-adc-debounce must send
> > correct coordinates all the time, especially in the case of pen release
> > when he knows that these coordinates are *clearly incorrect*.
> 
> > Note that an event is still generated on pen release, but it has no 
> > coordinates.
> 
>         Is it quote from Documentation/*, or from
> drivers/input/touchscreen/*.c in mainline? Because otherwise, it's
> nice, but still random, guess. And let's do it right now, not
> flip-flop it the way someone sees fit for this moment.

I just opened random touchscreen driver, like
drivers/input/touchscreen/touchwin.c

And they do not send x/y coordinates on untouch event.

> > - Damien Tournoud
> 
> -- 
> Best regards,
>  Paul                            mailto:[EMAIL PROTECTED]

Good day,

-- 
Anton Vorontsov
email: [EMAIL PROTECTED]
backup email: [EMAIL PROTECTED]
irc://irc.freenode.org/bd2
_______________________________________________
Hx4700-port mailing list
[email protected]
https://www.handhelds.org/mailman/listinfo/hx4700-port

Reply via email to