Always exec
Any automation project more complex than a simple relay request features interfaces and subsystems that independently store the unit states and keep them in sync with the real hardware equipment. But sometimes synchronization is either impossible or failed; occasionally the state updating is not specified in the configuration at all. In some cases, we may not even know whether the relay went off or not. A typical reason is unreliability of the relay and its failure to return the current status.
In this case, there is a “foolproof”: the action which is always executed, even if the unit state is the same as requested and is not supposed to be changed.
It is possible in the following cases:
If it’s a simple electrical circuit without any inner logic in the script: for example, a lamp or an outlet. If the lamp has already been turned on, nothing will change if the relay will be switched on - it may itself discard such a command. But both us and our system will be sure it’s really on.
It is not recommended in the following cases:
If it’s a complex control script, such as the door opener. Once the script is executed, we will switch on the motors and try to open the door. If it’s already open, you can only rely on the stoppers in the motor controller if any. If there are none - the motor will keep moving until the mechanism breaks down. Of course, the motor controller is always equipped with stoppers, but it’s better to keep it safe than be sorry.
It is impossible in the following cases:
If you run the actions “on” and “off” in the same way: e.g. to switch on a TV-set by sending a signal to IR-controller. A TV-set often has same button (and IR signal) for “on” and “off”. As a result, we will think that a TV-set is “switched on again”, but it is switched off instead.
If you run a command that can damage your equipment or cause an undesirable effect when “on” is called for the unit which is already “on”: the example can be door or motor without any stoppers.