« on: September 10, 2007, 07:52:23 am »
That's interesting, I wasn't able to find device support for that RCS anywhere in the source.
Maybe it was a version 1 thing?
I somehow don't think that info is accurate. The only X10 support I see at all is the CM11A. The CM11A code only has a small bit of support for receiving of X10 events. In fact, how CM11A handles received events demonstrates exactly what I feel is a problem with Pluto's current x10 support. There's too much "higher level" logic going on, its mapping events back to their LCME devices and firing "Sensor tripped" events. We should be able to implement a driver that strictly sends and receives native X10 and let another layer in the framework support the various types of X10 devices that one might control via X10. When you get into the various types of dimming wall switches and scenario setups, there's really quite a lot of logic that should not have to be re-done on a per-controller basis.
If I had to wager a guess, I'd say Pluto was rightfully focusing on newer, better technologies like Z-wave for automation. But in the spirit of hobbyist projects I think X10 support shouldn't be overlooked. In about 20 minutes I was able to install Homeseer on a windows box and get it controlling my Powerlinc USB, my RCS thermostat and all my lights. Homeseer is just too "beta" feeling for me to want to pay their ridiculously high prices when I can contribute my time towrds LinuxMCE.
Thanks for the info regarding the Z-wave thermostat. If I can track down which commands & events its using that will be helpful. There's an aprilaire interface I was looking at but it seems incomplete.
As for what I'm doing Project Wish is pretty easy to integrate. I'm more concerned with how to implement it in a way that's "friendly" to say, your Powerlinc V2 interface if/when you add X10 support to it. It doesn't look like Wish supports V2, and it doesn't look like iLink supports V1. Both Powerlincs use HID (and in fact, even have the same errata under Windows XP, breaking DirectInput support). So while I'm sure there was some opportunity for an integrated driver solution, we seem to be better off using different low layer approaches.