DAX interpreter BUG!!!

TBH what you are saying makes no sense to me:  variables 

are evaluated in the context in which they are written.

Variables should be visible in the scope in which they are declared, period.  Are you saying that DAX scoping isn't following what we are used to with all major programming languages out there?  

In any case, all my variables are usually declared top level, ie right after the 'MeasureName = ' line.  Therefore I would expect this scope to be visible all the way down to the RETURN statement.

Isn't it so?

One more thing... I read RADACAD's explanation, however it does not apply here.  Look again carefully at the code I posted:

"DownTime", IF ( [ErrorCategory] == "Unreachable",
                         VAR __unreachable_dt = DATEVALUE([UnreachableSince]) + TIMEVALUE([UnreachableSince])
                         VAR __unreachable_dt_in_secs = DATEDIFF(__start_date, __unreachable_dt, SECOND)
                         RETURN
                               IF( __unreachable_dt_in_secs >= 0,
                                     DATEDIFF(DATEVALUE([UnreachableSince]) + TIMEVALUE([UnreachableSince]), [ClosedDT], SECOND),
                                     BLANK() //DATEDIFF([CreationDT], [ClosedDT], SECOND)
                                ),

__unreachable_dt is declared right before the RETURN statement and when referenced in the RETURN block inside de DATEDIFF() inside the IF() as the 2nd parameter, the result is not that the computation is wrong.  

In other words, the block of code is

IF( condition,
    VAR __declaration0 = ...
    VAR __declaration1 = ...
    
    RETURN
        IF( __declaration1 >= 0,  // no issues here referencing the second variable 
            DATEDIFF( __declaration0, ... ) // issue here CAN'T FIND THE VARIABLE!!!
...

Is this really as it should be???

The result is that the interpreter complains the variable does not exist or is not defined!  Two very different things, yes?

Hello, any updates?