Dekker1/dekker-phd-thesis
Archived
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Yesterdays work on the conclusions

Browse Source
This commit is contained in:
Jip J. Dekker 2021-06-23 11:06:44 +10:00
parent 9c53a8a1b6
commit 8a73209de6
No known key found for this signature in database
GPG Key ID: 517DF4A00618C9C3
1 changed files with 36 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
chapters/6_conclusions.tex
View File

@ -17,9 +17,43 @@ Crucially, the proposed improvements in the performance of the rewriting process
This chapter presents the conclusions of this thesis.
We present a summary of the research and its contributions and discuss the future work arising from them.
\paragraph{Architecture}
\paragraph{Rewriting Architecture} In \cref{ch:rewriting}, we presented the principle contribution of this thesis: \emph{an architectural design for the rewriting of a high-level \cml{}}.
To achieve this we introduced an execution framework for \microzinc{}, a minimal \cml{}.
This framework offers the first \jip{something}.
A complex \cml{}, such as \minizinc{}, can be reduced to \microzinc{} and as such enjoy the same guarantees \jip{blah}.
\paragraph{Half Reification}
The framework operates on \nanozinc{}, a language that expresses (partial) \solver{} specifications.
Distinctively, \nanozinc{} has the ability to bind \constraints{} that define a \variable{} to that \variable{}.
This ensures that if it is discovered that a \variable{} is no longer required (\ie{} it is no longer referred to by any \constraints{}), it isn't kept in the model for the \constraints{} that define it.
Crucially, the framework is easily extended with well-known techniques to improve the quality of the produced \solver{} specifications.
\begin{itemize}
\item We described how \constraint{} \gls{propagation} can actively simplify the both \constraints{} and the \glspl{domain} of \variables{} during the rewriting process.
\item \gls{cse}
\item \constraint{} \gls{aggregation}
\item \gls{del-rew}
\end{itemize}
Two prototype programs were developed to evaluate this architecture:
\begin{itemize}
\item a compiler that translates \minizinc{} models to a \microzinc{} bytecode
\item a \microzinc{} byte-code interpreter that produces \nanozinc{} specifications
\end{itemize}
\jip{This should probably say something about its performance}
This framework enables many avenues of further research.
For one, the formal rewriting rules presented open up possibilities to more extended formal reasoning about \cmls{}.
This could potentially lead to the ability to proof certain properties of the rewriting process.
We introduce
\paragraph{Half Reification} \Cref{ch:half-reif} we further the
\paragraph{Incremental}