From: <ko...@us...> - 2000-03-20 18:08:35
|
I hoped you enjoyed the trip. Unfortunately I didn't get the time I was hoping for this past weekend to test, but I did manage do to the following: -noticed the problem is easily reproduced (for me) if I hold down a button on a PalmPad. PalmPads seem to generate more of these errors in the first place -the problem doesn't occur as much (hardly at all) when using the remote that comes with the MP3 Remote (different than the MouseRemote). I noticed that no matter how long I hold down one of its buttons, its led always lights for the same short duration. -ran mh with the X10 debug option. I noticed something was being presented to mh at every button press, but mh wasn't able to understand the codes. - a 'good' reception of the B5 BON commands looked like this: --------------------------------------------------- CM11 data=90 hex=5a mh CM11 data=Z data_d=90 CM11 data=3 hex=03 receive buffer length=3, mask=00000010, data_h=0302e1e2. CM11 db: data=B5BJ --------------------------------------------------- - a not so good reception (if I hold the button for a 2 second count) looks like this. mh saw the command but saw some other junk too. --------------------------------------------------- CM11 data=90 hex=5a mh CM11 data=Z data_d=90 CM11 data=5 hex=05 receive buffer length=5, mask=00001010, data_h=050ae9e2e9e2. CM11 db: data=B6BJ X10 db: try breaking XB6BJe9e2 into 4 and 2 character chunks: Found match: XB6BJ X10 db: try breaking XB6BJe9e2 into 4 and 2 character chunks: No match found: Xe9 X10 db: try breaking XB6BJe9e2 into 4 and 2 character chunks: No match found: Xe2 Running serial_data_buffer string: XB6BJ Serial data: X10 merged string=XB6BJe9e2B6BJ --------------------------------------------------- - and this one worried me even more. --------------------------------------------------- CM11 data=90 hex=5a mh CM11 data=Z data_d=90 CM11 data=5 hex=05 mh CM11 data= áâáâ data_d=5 --------------------------------------------------- Not only did mh not understand that I pressed a PalmPad button, but it looks as though the debug message is corrupt. Does Perl allow for programs to scribble on them selves, like an index error while accessing an array? Like I said, in all cases a light set to the same House/Unit code acted properly, even though mh didn't see the commands. I'm going to do more digging and will report what I find, any ideas/hints are always welcome, thanks. Kevin "Bruce Winter" <br...@mi...>@lists.sourceforge.net on 03/19/2000 03:56:21 PM Please respond to mis...@li... Sent by: mis...@li... To: <mis...@li...> cc: Subject: RE: [misterhouse-users] lost/corrupted signals Sorry for the delay. I just got back from a week in Montana. > Some Questions: > -Does the speed/os of the mh host make a significant difference to the > serial code mh uses to interface with the CM11A? Nope, it shouldn't. With the relativly slow serial interface and a buffer, this should not be an issue. If it were, I think we would see 'buffer overrun' errors in the console window. > -Does the CM11A report constructed ASCII strings/commands through the > serial interface, or is there some (timing related?) mh code that > constructs the received commands sequences? mh will try to string characters together within a 5 second window (controled by mh.ini parm x10_multikey_delay) to see if they match any defined serial item. That allows us to accept treat A1AJ (device A1 On, as we might get from a one button palmpad keypress) the same as A1 .... AJ (A1, then seconds later, AJ/ON, as we might get from a 2 button sequence x10 control keypad). > -How does the TW523/some_controller combination compare to a CM11A? And > what controllers have people tried (Weeder, Lynx10, CPU-XA, JDS, ......) The TW523 does the raw X10 protocol, so requires finer timing control than the CM11, which accepts higher level commands. I used to use the Weeder kit, but now use the CM11, as it can do extended data. > -is the supported hardware list up to date, or stated another way, does mh > support the Ocelot? Not yet, but someone else just sent me private email with that question (Ocelot support). I don't know what the pros/cons of it are or how well it's protocol is docuemnted. ________________________________________________________ To unsubscribe from this list, go to: http://sourceforge.net/mail/?group_id=1365 |