Changes in / [1b1a8da:4cdfcbd]

Location:

doc/theses/aaron_moss_PhD/phd

Files:

• experiments.tex (modified) (1 diff)
• introduction.tex (modified) (2 diffs)

doc/theses/aaron_moss_PhD/phd/experiments.tex

@@ -63 +63 @@
 The GC could be used for memory management with few changes to the code-base, but without a substantial re-write to enforce the same ``!const! children'' discipline \CFACC{} could not take advantage of the potential to share sub-objects; without sharing of sub-objects the GC variant of \CFACC{} must do all the same allocations and deletions and garbage-collector overhead degraded performance unacceptably (though it did fix some known memory leaks introduced by failures of the existing manual memory management scheme).
 
-Another minor architectural difference between \CFACC{} and the prototype system is that \CFACC{} makes extensive use of the pointer-chasing !std::list!, !std::set!, and !std::map! data structures, while the prototype uses the array-based !std::vector! and the hash-based !unordered_! variants of !set! and !map! instead. 
-Work is ongoing to port \CFACC{} to use these more efficient data structures.
-% TODO see how Thierry gets on with this
+Another minor architectural difference between \CFACC{} and the prototype system is that \CFACC{} makes extensive use of the pointer-based !std::list!, !std::set!, and !std::map! data structures, while the prototype uses the array-based !std::vector! and the hash-based !unordered_! variants of !set! and !map! instead. 
+Porting the prototype to use the pointer-based data structures resulted in modest performance regressions, whereas preliminary results results from porting \CFACC{} to use !std::vector! over !std::list! also showed performance regressions, in some cases significant.
+The relative performance impact of this architectural difference is unclear, and thus excluded from consideration.
 
 The final difference between \CFACC{} and the resolver prototype is that, as an experiment in language usability, the prototype performs resolution-based rather than unification-based assertion satisfaction, as discussed in Section~\ref{resn-conclusion-sec}. 

doc/theses/aaron_moss_PhD/phd/introduction.tex

@@ -2 +2 @@
 
 The C programming language has had a wide-ranging impact on the design of software and programming languages. 
-In the 30 years since its first standardization, it has consistently been one of the most popular programming languages, with millions of lines of C code still in active use, and tens of thousands of trained programmers producing it. The TIOBE index\cite{TIOBE} tracks popularity of programming languages over time, and C has never dropped below second place:
+In the 30 years since its first standardization, it has consistently been one of the most popular programming languages, with billions of lines of C code still in active use, and tens of thousands of trained programmers producing it. The TIOBE index\cite{TIOBE} tracks popularity of programming languages over time, and C has never dropped below second place:
 
 \begin{table}[h]
@@ -18 +18 @@
 \end{table}
 
-The impact of C on programming language design is also obvious from Table~\ref{tiobe-table}; with the exception of Python, all of the top five languages use C-like syntax and procedural control structures. 
-\CC{} is even a largely backwards-compatible extension of C, with development dating back nearly as far as C itself. 
-Though its lasting popularity and wide impact on programming language design point to the continued relevance of C, they also highlight the widespread desire of programmers for languages with more expressive power and programmer-friendly features; accommodating both low-impact maintenance of legacy C code and low-effort development of the software of the future is a difficult task for a single programming language.
+The impact of C on programming language design is also obvious from Table~\ref{tiobe-table}; with the exception of Python, all of the top five languages use C-like syntax and control structures. 
+\CC{} is even a largely backwards-compatible extension of C. 
+Though its lasting popularity and wide impact on programming language design point to the continued relevance of C, there is also widespread desire of programmers for languages with more expressive power and programmer-friendly features; accommodating both maintenance of legacy C code and development of the software of the future is a difficult task for a single programming language.
 
-\CFA{}\footnote{Pronounced ``C-for-all'', and written \CFA{} or \CFL{}.} is an evolutionary modernization of the C programming language which aims to fulfill both these ends well. 
+\CFA{}\footnote{Pronounced ``C-for-all'', and written \CFA{} or \CFL{}.} is an evolutionary modernization of the C programming language that aims to fulfill both these ends well. 
 \CFA{} both fixes existing design problems and adds multiple new features to C, including name overloading, user-defined operators, parametric-polymorphic routines, and type constructors and destructors, among others. 
 The new features make \CFA{} more powerful and expressive than C, while maintaining strong backward-compatibility with both C code and the procedural paradigm expected by C programmers. 
+Unlike other popular C extensions like \CC{} and Objective-C, \CFA{} adds modern features to C without imposing an object-oriented paradigm to use them.
 However, these new features do impose a compile-time cost, particularly in the expression resolver, which must evaluate the typing rules of a significantly more complex type-system.
 
 This thesis is focused on making \CFA{} a more powerful and expressive language, both by adding new features to the \CFA{} type system and ensuring that both added and existing features can be efficiently implemented in \CFACC{}, the \CFA{} reference compiler. 
-Particular contributions of this work include design and implementation of 
-parametric-polymorphic (``generic'') types in a manner compatible with the existing polymorphism design of \CFA{} (Chapter~\ref{generic-chap}), a type environment data structure based on a novel variant of the union-find algorithm (Chapter~\ref{env-chap}), and a new expression resolution algorithm designed to quickly locate the optimal declarations for a \CFA{} expression (Chapter~\ref{resolution-chap}). 
-This expression resolution algorithm was designed with the aid of a purpose-built prototype system which encapsulates the essential aspects of the \CFA{} type system without incurring the technical debt of the existing system or the friction-inducing necessity of maintaining a working compiler; the goal of this prototype system was to discover effective heuristics to avoid performing unnecessary work in the process of locating the optimal \CFA{} expression resolution. 
-This prototype system (described in Chapter~\ref{expr-chap}) shows promise as a vehicle for future research in compiler design, and is itself a technical contribution of this thesis.
+Particular contributions of this work include:
+\begin{itemize}
+\item design and implementation of parametric-polymorphic (``generic'') types in a manner compatible with the existing polymorphism design of \CFA{} (Chapter~\ref{generic-chap}),
+\item a new expression resolution algorithm designed to quickly locate the optimal declarations for a \CFA{} expression (Chapter~\ref{resolution-chap}),
+\item a type environment data structure based on a novel variant of the union-find algorithm (Chapter~\ref{env-chap}),
+\item and as a technical contribution, a prototype system for compiler algorithm development which encapsulates the essential aspects of the \CFA{} type system without incurring the technical debt of the existing system or the friction-inducing necessity of maintaining a working compiler (Chapter~\ref{expr-chap}).
+\end{itemize}
 
-Though the direction and validation of this work was fairly narrowly focused on the \CFA{} programming language, the tools used and results obtained should be of interest to a wider compiler and programming language design community. 
+Though the direction and validation of this work is fairly narrowly focused on the \CFA{} programming language, the tools used and results obtained should be of interest to a wider compiler and programming language design community. 
 In particular, with the addition of \emph{concepts} in \CCtwenty{} \cite{C++Concepts}, conforming \CC{} compilers must support a model of type assertions very similar to that in \CFA{}, and the algorithmic techniques used here may prove useful. 
-Type environments are also widely modelled in compiler implementations, particularly of functional languages, though also increasingly commonly in other languages (such as Rust) which perform type inference; the type environment presented here may be useful to those language implementers.
+Type environments are also widely modelled in compiler implementations, particularly for functional languages, though also increasingly commonly for other languages (such as Rust) which perform type inference; the type environment presented here may be useful to those language implementers.