hi,
Am Donnerstag 03 März 2005 16:48 schrieb Georg Holzmann:
Hallo!
Unless you want to be an alpha tester for the [hid] object (which would be much appreciated :), I would suggest using [linuxevent] for now.
Of course I want to be an alpha-tester :) - but I tried it and it didn't work. If I press a key, it is detected, but not which key ... (I tried it with the hid version from your hp) I also tried [linuxevent] and there is the same problem.
hmm, even linuxevent doesnt report them ? altough i know that it is a different object than the ev_dev i once wrote for jmax, i am pretty sure that at least with the event-interface you have to get the values.
for example, pressing the key "a" generates a event of type 1 (key event), with a id of 30 (thats the id for the key), and a 1 as event value (=pressed). if the auto-repeat snaps in, the event-value will become 2. on release, it becomes 0 .... key "s" is id 31, btw, so the id's are more or less orientated by the physical arrangement of the keys, instead of the alphabet ....
there might be a trap with the event system you might be falling into .... on newer kernels, the event system also outputs event of type "0", which i suppose is some kind of sync, or idle indication, i dont know. maybe you are not filtering out the "0" event ...... if i dont do it here, i always see only that 0 event, because it is sent after each other event ...
greets,
chris