Invalid example of variability error

[henrikt-ma]

The example in 3.8.3 Discrete-Time Expressions contains the following false statement about violating variability rules:

equation
  b = noEvent(x > 1) // error, since b is a discrete-time expr. and
                     // noEvent(x > 1) is not a discrete-time expr.

The problem is that there is no stated rule that allows use of variability to analyze normal equations. All one can say is that the equation cannot be used to solve for anything, but with the current specification, one needs to draw that conclusion without help of expression variability rules. Instead, one must go to the definition of variability of a variable to see that this equation cannot be used to solve for b, and then it becomes evident that the equation is useless and cannot be solved.

[henrikt-ma]

A remedy for this situation, that will allow allow this equation to be rejected based on the perfect matching rule and variability analysis, is proposed in the following PR:
#2526