[INSTEON] State of PLM scenes is not being kept

[peloy]

Activating a PLM scene works well, i.e. the devices in the scene are set to the right state, but apparently the internal state of the PLM scene itself is not set. Here's an example:

1. $upstairs_hallway_light is the PLM scene.
2. $upstairs_hallway_light controls $upstairs_hallway_slav1_light, $upstairs_hallway_slav2_light, and $upstairs_hallway_mast_light (this is a multi-way circuit where one light can be controlled from three switches).

Activating the PLM scene results in this:

05/05/13 12:40:16 PM Ran telnet command: upstairs hallway light on
05/05/13 12:40:17 PM [Insteon_PLM] DEBUG2: Sending obj=$upstairs_hallway_light; command=on; extra=FF incurred delay of 0.00 seconds; starting hop-count: 3
05/05/13 12:40:17 PM [Insteon::BaseObject] $upstairs_hallway_light::set(on, $upstairs_hallway_light_v)
05/05/13 12:40:17 PM upstairs_hallway_slav1_light newly set to on set by $upstairs_hallway_light
05/05/13 12:40:17 PM upstairs_hallway_mast_light newly set to on set by $upstairs_hallway_light
05/05/13 12:40:17 PM upstairs_hallway_slav2_light newly set to on set by $upstairs_hallway_light
05/05/13 12:40:18 PM [Insteon_PLM] Received all-link cleanup success: obj=$upstairs_hallway_light; command=on; extra=FF

However, the state of $upstairs_hallway_light (the PLM scene) is not set... shouldn't that last message received from the PLM trigger a state update?

To confirm, I put in the following user code, which never runs:

if ($state = $upstairs_hallway_light->state_now() ) {
print("upstairs_hallway_light changed state to $state.\n");
}

A consequence of this is that in the web interface, PLM scenes are always displayed in the same state and cannot be changed through buttons in the Lights/Appliances screens.

Looking in the web interface at the state log for this particular scene shows that the state was being set in the past, including set_by, but there are no recent events, which leads me to think that a recent change caused a regression.

Running mstovenour's merge_i2_aldb_support branch.

[krkeegan]

I worked on the PLM Scenes, and it is quite possible that I broke this. I don't imagine it will be too difficult to track it down and fix it.

[krkeegan]

OK, the above patch fixes this. It is based off of the i2_merge, so we need that to take place first.

[peloy]

On 05/07/2013 09:43 PM, krkeegan wrote:

OK, the above patch fixes this. It is based off of the i2_merge, so we
need that to take place first.

Thanks! I am running with this patch now. Will report back if I see any
issues.

Cheers,

Eloy Paris.-