NAG-003185 · Historical whole-proof gap

Variation control and convergence against continuous multipliers

Records without a chapter binding

Remaining historical proof gap

The sealed ledger records a gap in the proof of the historical whole statement.

Historical statement · exact TeX
\begin{theorem}[Variation control and convergence against continuous multipliers]\label{f9tf:convergence}
Work on a fixed local decomposition \eqref{eq:decomp}. For every fixed compact test $\beta\in\E^{n-1,n-1}(U)$, the measures $\nu_{u_\epsilon,\beta}$ have common compact support in $\supp\beta$, uniformly bounded total variation for $0<\epsilon\leq1$, and converge to $\nu_{u,\beta}$ against every continuous scalar on that compact set. On each fixed cutoff stage the variation bound is uniform in $\epsilon$ with coefficient order zero.

More explicitly, for a stage domination form $\theta_i$ of Theorem~\ref{f9tf:measure}, one may take
\begin{equation}\label{eq:uniformmass}
 A_{u,i}=c_i\sum_a|q_a|
 \left(T_{w_a}(\theta_i)+\log m_a\,|\mu_{U,\DD\theta_i}|(U)\right),
\end{equation}
with $\log m_a=0$ for an unchanged smooth convex piece. Then
$|\nu_{u_\epsilon,Q_i\eta}|(U)\leq A_{u,i}p_{t_f(V),0}(\eta)$.
For an arbitrary fixed $\beta$, a finite decomposition into cutoff images gives a bound $M_{u,\beta}$ for these variations and the variation of $\nu_{u,\beta}$.

If $v_\epsilon\to v$ uniformly on $\supp\beta$, then
\begin{equation}\label{eq:joint}
 \int v_\epsilon\,d\nu_{u_\epsilon,\beta}
 \longrightarrow\int v\,d\nu_{u,\beta}.
\end{equation}
The result also holds for an ordinary smooth $v$ left unchanged. It is proved for this fixed positive decomposition and its explicit approximants; mere weak convergence of arbitrary signed currents is not substituted for these hypotheses.
\end{theorem}
Reader cross-references

This open step in the proof map