Hi list,
I am using [hidraw] to read events from a mouse. I understand that hidraw in fact reports the entire state of the device rather whenever some event happens. Interestingly, moving the scroll wheel down by one step is reported, but moving it down another step afterwards is not reported unless a different event (say, moving the mouse a pixel) is done. It makes somehow sense if I think of the mouse wheel as a state device, where a repeated "wheel down" is not considered as a change of state.
Nevertheless it seems impossible to parse repeated wheel down actions. Does anyone have an idea (besides using [hid] instead)?
Thanks! Peter
there could be a bug in [hidraw]'s internal change mechanism. (missing some change?)
not sure when i'll try to debug.
I had created an issue at https://github.com/Lucarda/pd-hidraw/issues/4
--
Mensaje telepatico asistido por maquinas.
On 16/04/2024 06:55, Peter P. wrote:
Hi list,
I am using [hidraw] to read events from a mouse. I understand that hidraw in fact reports the entire state of the device rather whenever some event happens. Interestingly, moving the scroll wheel down by one step is reported, but moving it down another step afterwards is not reported unless a different event (say, moving the mouse a pixel) is done. It makes somehow sense if I think of the mouse wheel as a state device, where a repeated "wheel down" is not considered as a change of state.
Nevertheless it seems impossible to parse repeated wheel down actions. Does anyone have an idea (besides using [hid] instead)?
Thanks! Peter
@Peter,
yes, there's a bug. [hidraw] should not have an internal change mechanism.
i had used your mouseweel example to debug.
I should fix this soon with a new version and also upgrade to hidapi 0.14.0
thanks,
:)
--
Mensaje telepatico asistido por maquinas.
On 16/04/2024 20:31, Lucas Cordiviola wrote:
there could be a bug in [hidraw]'s internal change mechanism. (missing some change?)
not sure when i'll try to debug.
I had created an issue at https://github.com/Lucarda/pd-hidraw/issues/4
--
Mensaje telepatico asistido por maquinas.
On 16/04/2024 06:55, Peter P. wrote:
Hi list,
I am using [hidraw] to read events from a mouse. I understand that hidraw in fact reports the entire state of the device rather whenever some event happens. Interestingly, moving the scroll wheel down by one step is reported, but moving it down another step afterwards is not reported unless a different event (say, moving the mouse a pixel) is done. It makes somehow sense if I think of the mouse wheel as a state device, where a repeated "wheel down" is not considered as a change of state.
Nevertheless it seems impossible to parse repeated wheel down actions. Does anyone have an idea (besides using [hid] instead)?
Thanks! Peter
@Peter
v0.1.1 is up on deken: https://deken.puredata.info/results.html?name=hidraw
on v0.1.0 the subsequent mouseweel scroll (more precisely all the message) was interrupted because it was the same as the previous message. Now on v0.1.1 the message gets transmitted but is the same as the previous. Nothing gets transmitted if the second outlet does not output 2 (read Ok).
--
Mensaje telepatico asistido por maquinas.
On 17/04/2024 00:06, Lucas Cordiviola wrote:
@Peter,
yes, there's a bug. [hidraw] should not have an internal change mechanism.
i had used your mouseweel example to debug.
I should fix this soon with a new version and also upgrade to hidapi 0.14.0
thanks,
:)
--
Mensaje telepatico asistido por maquinas.
On 16/04/2024 20:31, Lucas Cordiviola wrote:
there could be a bug in [hidraw]'s internal change mechanism. (missing some change?)
not sure when i'll try to debug.
I had created an issue at https://github.com/Lucarda/pd-hidraw/issues/4
--
Mensaje telepatico asistido por maquinas.
On 16/04/2024 06:55, Peter P. wrote:
Hi list,
I am using [hidraw] to read events from a mouse. I understand that hidraw in fact reports the entire state of the device rather whenever some event happens. Interestingly, moving the scroll wheel down by one step is reported, but moving it down another step afterwards is not reported unless a different event (say, moving the mouse a pixel) is done. It makes somehow sense if I think of the mouse wheel as a state device, where a repeated "wheel down" is not considered as a change of state.
Nevertheless it seems impossible to parse repeated wheel down actions. Does anyone have an idea (besides using [hid] instead)?
Thanks! Peter
- Lucas Cordiviola lucarda27@hotmail.com [2024-04-17 09:26]:
@Peter
v0.1.1 is up on deken: https://deken.puredata.info/results.html?name=hidraw
Thank you for your quick response, fix and deken upload, Lucas!
The problem is still there with the current hidraw v0.1.1 and hidapi v0.14.0 on Linux that is, the same message (mouse wheel scroll down or up) does not get output (checked with print) unless I move the mouse one pixel in between.
Thanks again! Peter
On 20/04/2024 04:40, Peter P. wrote:
The problem is still there with the current hidraw v0.1.1 and hidapi v0.14.0 on Linux that is, the same message (mouse wheel scroll down or up) does not get output (checked with print) unless I move the mouse one pixel in between.
strange,
I can't replicate on Debian 12 with this connection:
[hidraw] | [print]
i get all scrolls reported (also all others mouse events)
PS: i'm using the latest deken pkg on a live distro usb stick.
--
Mensaje telepatico asistido por maquinas.
-
Lucas Cordiviola -
Peter P.