From: Bruce W. <br...@mi...> - 2001-02-10 19:25:40
|
> My pleasure, I can see why that would happen now (technically, it's acting > just like I designed it). Bruce, do you see this as correct behaviour or > should we filter out 'manual' commands? (My guess is no, but...) > > Second, should MH even be generating the manual if the full string was > received within the timeout for putting strings together? That would tend > to indicate for the SmartLinc triple on that MH would generate states of: > manual, on, on, double on, on, triple on for one triple on case. > I do think that would be the wrong behaviour. If the X10 data comes in all at once, then we get just one state out of it. If, however, it comes in in chunks, then we get any state that matches whatever data we matches as it comes in. For example: XA1AJ -> only matches ON XA1 ... XAJ -> matches manual first, then ON XA1AJAJ -> only matches double ON XA1AJ ... XAJ -> matches ON, then double ON We could try to put a delay in there, but the size of that delay might be debatable, and it would slow down the detection of other things. I think we are better of ignoring states we don't care about. Bruce |