why the objective function of the SQP subproblem is not the second-order Taylor expansion of f in the general case
-->
why the SQP subproblem should not minimize second-order Taylor expansion of the original objective function unless the constraints are all linear.
https://github.com/ragonneau/phd-thesis/blob/f5e9029df3534f5d076922ead00db04702703cfb/content/sqp.tex#L27
why the objective function of the SQP subproblem is not the second-order Taylor expansion of f in the general case --> why the SQP subproblem should not minimize second-order Taylor expansion of the original objective function unless the constraints are all linear.