On 02/26/2010 10:44 PM, Jarod Wilson wrote:
> That all looks perfectly normal, they're clearly labeled as repeats.
> (the incrementing value in column two). Just suppress repeats in your
> lircrc and/or run lircd with the -R (--repeat-max) X option to limit
> things to X repeats. I use the lircrc repeat suppression myself.
OK then. I used repeat limits with that remote anyway, it was awfully
unpredictable. Looks like these buttons are a bit dodgy too. :) I just
thought this kind of behaviour could be smoothed out by a lircd config.
Granted, physically the panel is really sending all those events so it's
not really an issue of LIRC.
Anyway, thanks for those tips!
BTW, I had to make a tiny patch for lircrcd, it was segfaulting on me
because of unhandled NULL dereference in CODE handler (triggered when
client didn't send \n). I'm using kind of complex modes and really
needed the daemon:
RCS file: /cvsroot/lirc/lirc/tools/lircrcd.c,v
retrieving revision 5.4
diff -r5.4 lircrcd.c
> if(arguments == NULL)
> return send_error(fd, message, "protocol error\n");
The source application (mythtv) was using LIRC "template" lirc_client.c,
the newline character is still missing in CVS. I talked about it with
them they accepted it as a bug, but asked me to let you know to fix this
"in the source":
RCS file: /cvsroot/lirc/lirc/tools/lirc_client.c,v
retrieving revision 5.29
diff -r5.29 lirc_client.c
< sprintf(command, "CODE %s", code);
> sprintf(command, "CODE %s\n", code);
Here's my original post in mythtv-dev: