Based on the suggestions that I received from all you kind folks, I have worked out a mechanism that I think may work which uses toadtogs to track and switch the input states of my TV (going back to my very original post on this thread). Details follow - please tell me if I am totally "out-to-lunch" or if this is a sensible approach:
1. Assign one toadtog bit for every state to be tracked. The "on" toadtog is used to identify the "active" input state.
2. Assign four families of keypresses to track and toggle states and to send the actual IR commands.
3. Assign the four keypresses for every toadtog bit (or state to be tracked) as follows:
3.1 Keypress is: shift-<number> on device=TV which means "Switch TV to Input <number>"
3.1.1 underlying toadtog is ForceOn(bit<number>)
3.1.2 if originally off do:<family 2 keypress for input <number+1>> to start search for previous input state.
3.1.3 if originally on do: nothing (the correct input state is already setup)
3.2 Keypress is: <family 2 keypress for input <n>> device=TV which is used to search for the previous input state.
3.2.1 underlying toadtog is TestOnly(bit<n>)
3.2.2 if off: <family 2 keypress for input <n+1>> to continue searching for previous input state
3.2.3 if on do: <family 3 keypress for input <n>>
3.3 Keypress is: <family 3 keypress for input <n>> device=TV which is used to deactivate previous input state.
3.3.1 underlying toadtog is ForceOff(bit<n>))
3.3.2 if originally off: nothing (this should never happen but if it does the processing simply stops)
3.3.3 if originally on: <switch to this device and send keypresses (such as discrete power-off) to deactive this input>, <switch back to TV>, <send family 4 keypress for input<n>>
3.4 Keypress is: <family 4 keypress for input <n>> device=TV which is used to send "next video-input" IR command to TV until selected input state is reached
3.4.1 underlying toadtog is TestOnly(bit<n>))
3.4.2 if originally off: <next video-input keypress>, <pause keypress>, <family 4 keypress for input<n+1>>
3.4.3 if originally on: <switch to this device and send keypresses (such as discrete power-on) to activate this input> (end of processing)
My assumption is that the new toadtog state is set before any of the toadtog selected keypresses are executed. Can someone verify this?
This technique requires that one of the toadtogs must initially be switched on otherwise it goes into a loop. Something like shift-0 could be used to set a default state (toadtog bit) for the TV tuner input without doing anything else. The remote can synchronized to whatever TV input happens to be active by cupping your hand over the IR transmitter and pressing shift-<number> where the number corresponds to the active input.
To make this technique work for me, two things are needed:
- an additional 7 toadtog bits (I have already used four for discrete power on / off for each of four devices). This would bring the required number of toadtog bits to 11 and the documentation I have says I can only use 8. If I add another ToadTog device, do I get another 8 bits? How do I set this up in IR?
- more macro memory. The 28 DSMs for this program will exceed the amount of free macro memory. I have "scads" of upgrade memory available but only 224 bytes of keymove/macro memory left. Can the "partition" between keymove/macro memory and upgrade memory be moved?
I'm still open to other ideas as this technique consumes gobs of remote memory, and does not follow the KISS (Keep It Sweet and Simple) principle.
... Tony
Note I thought the same thing. No one had ever been so indiscreet as to write d i s c r e e t down for me so I didn't know that is was spelled differently.