clock-trigger: incompatibility with execution retry delays

[oliver-sanders]

The deprecated [special tasks]clock-trigger configuration appears to cause execution retry delays to be skipped.

With the following workflow the execution retry delays skipped, but if you swap out the clock-trigger for an xtrigger, then it behaves correctly:

[scheduling]                        
    initial cycle point = previous(T00)    
    final cycle point = PT1S        
    # [[xtriggers]]                                                                  
    #     clock = wall_clock(PT1H)    
    [[special tasks]]    
        clock-trigger = foo(PT1H)    
    [[graph]]    
        T00 = """    
            foo    
            # @clock => foo
        """                         
                                    
[runtime]                           
    [[foo]]                         
        execution retry delays = PT30S    
        script = false   

[MetRonnie]

When you say skipped, I have found it to do the specified retries, but immediately instead of waiting for the given period

[oliver-sanders]

I think that's probably what I meant.