Interesting... The XTB-232 mod could convert a %dim into a reduced resolution extended command as a mode option. Our Leviton switches accept extended commands, but not %dim.
Would this work with, or could it be retrofitted to, an unmodified XTB-232 already in use with X10 only (the CM11A protocol at 4800 bps)? Could the conversion occur in both directions (send: extcode to preset dim; receive: preset dim to extcode)?
I have several Insteon SwitchLinc dimmer switches still in use that use only the preset dim commands, which I have tried to cater to in my own code on the PC. Sending a preset dim command is the easy part and usually works fine. The problem, for me, is when the switches send their preset dim level back and I have to decode the virtual house code nonsense into usable information, using an asynchronous interface and scheduling scheme which could hardly be called "real time". I would be very happy if this problem were solved for me.
As I see it, there are a few problems with the preset dim arrangement of the (unofficial) protocol. But besides being weird and messy, the most significant one is that the sequence is not transactional; it is sent as 2 completely separate and independently valid commands. If you have line noise or interference by an "impolite" transmitter, whichever command was not corrupted now means something different. If you add collision detection or avoidance, the two commands could arrive intact but nonconsecutively; there could be 1 or more unrelated commands from another transmitter in the middle, again changing their meaning. If you're trying to send preset dim commands, you accordingly have to pay special attention to other activity on the line and resend the entire sequence if another transmission interferes. If you're trying to receive preset dim commands, the only thing you can do is try to sanitize your inputs, correlate the pieces together as best you can, and discard data that is obviously wrong.
I'm sure you realize all of this already. But I'm just trying to illustrate why this is a problem for me. And if you endeavor to decode incoming preset dim commands, this complexity may come with it. Unless I'm just totally missing something, which is quite possible!