\chapter{Conclusion}
\label{chap:conclusion}

In this work, we have considered approaches for automatically
adding support for concurrent updates to static data structures,
for the purpose of reducing the amount of work necessary to produce a
dynamic index. Classical dynamization techniques suffered from several
limitations on supported data structures, as well as performance problems
stemming from a lack of configurability and poor worst-case insertion
performance. We have attempted to address these limitations.

The result of these efforts is a generalized dynamization framework built
upon a set of novel mathematical results that allows for many static
data structures to be automatically extended with tunable, concurrent
insertion and deletion support, with bounded additional query cost. The
technique expands on the base Bentley-Saxe method with new query interfaces
to enable support for search problems that are not traditional decomposable,
a tunable design space including buffering and alternative block layout
polices to allow for trade-offs between insertion and query performance,
and support for parallel reconstructions in a manner that effectively
reduces the worst-case insertion cost while maintaining similar query
performance.