Discussion of "A robust method for equality constrained state estimation"

The paper's original author developed a new and valuable approach to state estimation modeling and solution. In several parts of the text, it was emphasized that an important benefit of the approach-perhaps the most important one, as could be inferred from the text-was that the gain matrix was positive definite, and so the use of mixed 1 x 1 and 2 x 2 pivots was not necessary in the factorization process. The author asks questions on this inference and the original author's response to these question is also shown.