Gerd Knorr wrote: > > It is possible to open /dev/input/event<n> and ask for exclusive > access via ioctl(). If you do that nobody else will see the events, > neither the linux console nor other applications which are reading > from the same device. Good. That solves part of the problem. > > But it's still unclear how one can configure the system to not put > > IR key events in the normal input queue. > > I don't want to disable it completely. It may be useful to have the > IR events in the linux console input queue. Maybe it is even better > than the current way to handle it because it just works with every > application, not only the ones with lirc support added. For example I > can browse teletext pages using mtt (or alevt) and control them using > the remote control. Neither mtt not alevtd has lirc support ... > Moving the cursor in emacs works too :) If you have an IR keyboard you need IR events in keyboard queue. But for a normal remote control it might be different, depending on what the user wants. The user must be able to configure. With just EVIOCGRAB you'd still need lircd to re-route events to applications according to config, and bypassing X focus. > This being configureable would be best, i.e. say "this (group of) > input device(s) should send keyboard events to the linux console". > Maybe even per virtual console, like it can be done with framebuffer > devices. I havn't found a way to do that through, I suspect it is > not implemented (yet?) :-/ Probably because there was no need for it so far. It'd best to discuss this on lkml and/or lirc-list. Johannes