B0: the frozen continuation

One clarification of XIV-A is recorded. Its Definition of ${\mathfrak{F}}_W$ described the fields other than $c$ as “unrestricted smooth sections”, while its $\mathfrak{F}$ had both ghosts compactly supported and its antifields compactly supported. Nothing in XIV-A used a noncompact antighost or a noncompact antifield; the breaking functional (eq:breaking) depends on $c$ alone, and the fermion pairing $\int \bar{c}\cdot (2D\gamma -b)$ is finite for compact $\bar{c}$. We therefore fix ${\mathfrak{F}}_W\subset \mathfrak{F}$ as in Definition def:spaces. The only consequence below is that the antighost sector of the free-shell intersection is also trivial (Theorem thm:ghost).

The negative freeze is stated in the same breath. XIV-B does not select an antifield-dependent observable completion, a correction of the lower coefficient $\kappa {\mathcal{A}}_1$, a spurion law for $g$ or ${\sigma }_{\epsilon }$, a source-switched upgrade, the convolution sheet, or a new integration interface outside ${\mathfrak{K}}_{\mathrm{loc},\mathrm{R}}$. Those were left outside XIV-A's class. Reusing the witness $u_{\ast }$ after changing the observable would be a new contract and is not done. The unrestricted master equation is not reopened: its failure is copied as (eq:witnessA1).