I seem to be having an is issue with the RM command for stop on truth change. With the following rule RM keeps repeating the action even after the condition becomes false.
Just noticed this myself as well.
I have a rule that will notify me if nobody is home and the garage door is open. However, even when it changed to false (the garage door was closed and somebody was home) it was still notifying me.
How I got around it is, was to put in the false a statement to stop the action on this rule and it seems to work for now. But I think something was either broken or changed.
I’m running the latest firmware 18.104.22.168.
Thanks, that will get me going again.
Yup. Same issue. Let me know if you need anything.
Your "workaround" seems to work fine, thanks for the tip.
I am not able to reproduce this failure. Could it be that the change in truth is happening during the delay period, before the repeat even starts?
The rule would be true for it to start. I then have a 15 minutes delay and then a repeat every 15 minutes.
During the 15 minutes delay, it would have switched to false. If I didn't do this then as I drive up into the driverway and open the garage door it would give me notifications as my presence hasn't yet been updated.
So during that 15 min delay it should change to false and cancel. But it wasn't cancelling at all and I kept getting notifications every 15 minutes after that even though it was false.
By adding in the false step to stop the actions on this rule it then would stop notifying me as expected.
It was working fine until the latest update. Its just not cancelling the true when it evaluates false.
It doesn't cancel the repeat, because the repeat hasn't happened yet.
What Delay these actions does is postpone all of the actions for the time specified. So repeating has not even started. Stop on truth change refers to the repeat, not to the delay before that. The repeat will stop when there is a truth change during the repeat period.
Make sense. After the delay does it reevaluate the condition before executing the action? Or will it first execute the action and then evaluate the condition again?
It was working before without me having to put in the false to stop it but the last update changed it. I'll review my other rules and update them as well if needed.
I agree, I've had my rules in place for quite some time, and didn't have to use the "false" workaround before either.
It only evaluates conditions when there are events that hit the conditions specified in the rule.
There haven't been any changes to how this stuff works in recent releases. When you say it was working before, what was working? It never would have stopped the repeat without a rule truth change happening after the repeat started. What about using cancel on truth change for the delay itself?
I think my issue was simply a misunderstanding of how the feature worked, and evidently I did not notice the fact that it didn't re-evaluate the condition in-between repeating.
When I say it was working before, I mean I didn't have to put in the cancel on the truth false. Once the false hit, the truth was automatically cancelled. I had to put the delay in because my presence wouldn't update before the garage opened, so I just wanted to delay the start of notifications. After adding that delay it would cancel on its own.
After the last update it wouldn't cancel the repeat, so yes the cancel was probably being executed during the delay.
But I have since added in the false part to cancel and it now works as it did before.
So all good now.
Was this a problem or a misunderstanding of the implied functionality?
Once I enter a repeat loop (count max of 10) if a truth change occurs, the repeat action continues???
Any insight sir?
Not seeing what you have set up, I can't really tell you.
There are major changes coming to this functionality in Rule 3.0, which is in beta test now and should be released soon. Rather than chase down oddness in Rule 2.5, I suggest you wait for this upcoming release.
This may sound like a strange request (as i'm somewhat inexperienced), but is rule 3.0 going to require that existing rules be written in any way? And if so, is there a general ballpark timeframe as to when "soon" is?
The reason I ask is I'm just a month into Hubitat and already have about 3+ dozen rules configured, come Thursday I will be having 2 more hubs arrive at which I'm setting up the coordinator hub (for rules, apps, etc.) a zwave hub and a zigbee hub.
Therefore I will already be writing these once from my current hub to the coordinator hub, so I will wait if they need to be re-written for 3.0 until it's released so I'm not doing the re-writing all of them twice.
You won't have to re-write any rules. Most of my rules are still the original Rule child, that pre-dates Rule 2.5.
New rules written after the update will be 3.0. Existing older rules keep on working, and can be edited as now.