Injective resolutions of \(BG\) and derived moduli spaces of local systems (Q1840480)

Classifying spaces for mathematical objects can often be constructed as algebraic stacks associated with group actions on finite connected CW-complexes and their corresponding locally constant sheaves of torsors. However, these algebraic stacks may not be smooth, which is due to the fact that jumping phenomena may occur in their first-order deformation theory and higher cohomology theory. It has been suggested by \textit{P. Deligne}, \textit{A. Drinfeld}, \textit{M. Kontsevich}, and others, on several occasions in the literature, that this difficulty might be overcome by working in the derived category instead, that is by constructing a nonabelian derived functor of the original moduli functor leading to a ``derived moduli space'' which is always smooth (in an appropriate sense) and whose tangent spaces involve the entire cohomology of the sheaf of infinitesimal automorphisms. The appropriate technique for constructing such derived moduli functors (in characteristic zero) seems to be the one of ``dg-schemes'', which are ordinary algebraic schemes equipped with a sheaf of (negatively graded) differential algebras. In the paper under review the author presents such a construction of a derived moduli space. As this program of constructing derived moduli functors via dg-schemes is highly non-straightforward and conceptually utmost involved, the author focuses on the ``simplest case'' of moduli of local systems. His construction is based upon the observation that those moduli spaces can be represented, topologically, as sets of simplicial homotopy classes of simplicial maps from a Čech nerve of a topological space \(X\) to the simplicial classifying space of the acting algebraic group \(G\). As for extending this set-up into a derived category, the author constructs an appropriate injective resolution of the classifying space \(BG\) of \(G\) and, via this resolution, suitable fibrations for dg-schemes. This leads to the building-up of the derived moduli stack of local systems, which appears as a dg-stack. The author's concluding main theorem states that the underlying space (derived moduli space) is a smooth dg-manifold with explicitly computable fundamental group and higher cohomology of the tangent dg-spaces. This result provides a ``derived'' generalization of the main theorem of \textit{V. Hinich} and \textit{V. Schechtman} on the local structure of moduli of local systems [cf.: ``Deformation theory and Lie algebra homology'', I=Algebra Colloq 4, No. 2, 213--240 (1997; Zbl 0905.17022); II=Algebra Colloq. 4, No. 3, 291--316 (1997; Zbl 0919.17014)]. As for the underlying theory of dg-schemes and dg-manifolds, the author provides the basic facts in the first part of the paper. The classical references for this part are given by the original papers of \textit{D. Quillen} [cf. ``On the (co-)homology of (commutative) rings'', Proc. Symp. Pure Math. 17, 65-87 (1970; Zbl 0234.18010)] and \textit{H. J. Munkholm} [cf. ``DGA algebras as a Quillen model category: Relation to shm maps'', J. Pure Appl. Algebra 13, 221--232 (1978; Zbl 0409.55018)]. Altogether, this is a very original and important paper in moduli theory.