The title says it all.
Could it be possible to use temp increase or decrease as conditions?
Say I set my trigger as Temp changed. Could I then have conditional actions based on whether the change is an increase or a decrease, and possibly, conditioned on the amount of increase or decrease?
I do exactly this for my water heater and furnace modes. If it becomes X degrees, (changed as trigger) I switch from heating to cooling mode, and similar but different temp to change back from cooling to heating.
For water heater, ifchanged below X temp, it is high demand mode or above X it goes to heat pump mode.
I would think this should be fairly easy to implement. Back when Stringify was a thing, I was able to do exactly what I am suggesting by, basically, creating a flow-chart in a simple GUI.
I would have done what you did above, or I would have used two sensors and compared ambient to stack temp.
I am out of my league much beyond this. There is an app you should check out for an example of how to do a rise, in this case, humidity. Smart Humidity Fan Maybe you could adapt that, or use it as an example for a new app. Please be sure to ask the author if you decide to publish any of this code though.
Why not at the end of your rule set the base temp to the current temp therefore at the next change you can check for the rise from the last time you ran the rule?
Could always use the WATO app, too. It has increasing/decreasing triggers. Although I don;t think you can have the rising a different threshhold than the decreasing. Anyway, take a look.
Since I have the blower turning off when temp decreases below 90, that would set the base to 90 and effectively my app becomes a simple thermostat turning on and off at 90.
.....nevermind, I got the logic now. The rule still runs whenever temp changes therefore, as temp drops off, the base is still being updated.
For my bathroom humidity fan control, I created a slow-moving average device, and then compare current humidity with the slow-moving average to detect when it rises. This has proven to be a very reliable detector of when someone is showering.
The averager App is derived from Bruces code, with some modifications by me.