The main purpose of this study was to review and modify some of the work that has been done concerning the generation of pseudo-random numbers. The PURGE3, MIX1, and MIX2 subroutines, written in FORTRAN IV to be run on the IBM 360 computer system, are used for the generation of pseudo-random numbers from a family of Pearson distributions as well as from any combination of mixtures of two Pearson distributions, a normal distribution with arbitrary mean and variance and a uniform distribution on any finite interval. The primary improvement undertaken here was to increase the efficiency of operation by enabling the PURGE3 and MIX1 subroutines to be called in a single computer program. The Pearson distribution may be specified either by the first four moments or from sample data. The parameters of the fitted distribution are printed and, if desired, the moments and a graph of the distribution of the 100 pseudo-random numbers generated are printed. A graph of the mixture of distributions generated in the MIX1 subroutine may be obtained from 10,000 random numbers generated from the mixture. The MIX2 subroutine functions to generate random numbers only. The speed of generation varies from about 13,000 random numbers per minute for a Pearson distribution which computes and prints the moments from the generated numbers, to approximately 143,000 numbers per minute if mixtures are used without computing the moments from the generated numbers. / Master of Science

This thesis describes a system that allows the designer of an interface between humans and computers to implement an English-like set of commands easily. Designers model commands on imperative English sentences containing verbs, direct and indirect objects, adverbs, adjectives, and prepositional phrases. The system restricts the ordering and presence command components as little as possible while still allowing the commands to be analyzed in a deterministic manner. Thus, commands formed with the system can mirror not only the form of English, but the flexibility of English as well. This system is based on the macro expansion of a command into a base language, and therefore has only the expressive power of the base language. Because of this, the system can only aid in the definition of commands, and not in the definition of a language. The language for describing the commands is structural in nature. The programmer defines the parts of a command instead of the syntax of the command. With each part of the command, the programmer defines sections of code that will be generated at various points during the recognition of the command. An implementation can execute generated code at some later time, at the end of the translation, or during the expansion. A prototype implementation expands of a file of commands into the VAX/VMS Digital Command Language. / M.S.

The thesis describes the requirement, design, and implementation of a software package that can be used to perform quantitative studies on certain aspects of a programming task. Of specific interest are experiments with the level of interactiveness of the human-computer interface relating to program construction and with language design principles relating to identifier scope rules. The software package for conducting these experiments is an interactive language environment called PEEP. Its base language is Pascal and its design is based on Johnston's semantic models of computation. Storage representations and implementation of these semantic models are described. These models depicting the compile-time structure, run-time structure, and realization of the static and the dynamic scoping rules. The evolution and current research of programming environments, user interface to PEEP, and future research on PEEP are also focuses of attention. / Master of Science

An expressive type language and the ability to do compile-time type inference are desirable goals in language design, but the attainment of the former may preclude the possibility of the latter. Specifically, the type conjunction operator (type intersection) induces a rich type language at the expense of decidability of the typeable expressions. Two extreme alternatives to this dilemma are to abandon type inference (and force the programmer to, essentially, supply a derivation for his type claims) or to abandon (or restrict) type conjunction. This work presents a third alternative in which the programmer, at times, may be required to supply explicit types in order for type inference to succeed. In this way, the power of conjunctive types is preserved, yet compile-time type inference can be done for a large class of polymorphic functions, including those typeable with parametric types. To this end, we introduce a simple combinator based language with typing rules based on type conjunction and a subtype relation, of sorts, called "weaker." The validity of the type rules with respect to the usual interpretation of "type" is shown, along with the undecidability of the type relation. It is shown how the computational portion of the language can be modified to accommodate explicit type information which may direct an automatic type derivation. This new language has the principal type property with respect to a decidable relation, although deciding this relation is shown to be an NP-Complete problem. The language is extended to accommodate type fixed points, and extended further to allow all expressions with parametric types to be typed automatically, and to accommodate integers, pairs, sums, and abstract types in the form of type generators. / Ph. D.

The research reported in this dissertation investigates extending the power and flexibility of an object-oriented language with inheritance, Static typing, and concurrency. A language supporting inheritance, static typing, and concurrency offers a significant leverage in software development. However, when these features are provided together, the benefits of the features are significantly reduced due to the interaction among them. The challenge for the designer of an object-oriented language lies in the difficulty of reconciling the conflicts among these features. This thesis discusses two issues: combining static typing with inheritance and combining concurrency with inheritance. A new model of type and inheritance, called HANA, is presented. The HANA model integrates multiple inheritance, multiple representation, method exclusion, and method name overloading with static typing. The contribution of HANA is that it extends other existing models of type and inheritance with enhanced expressive power, increased reusability, and improved program efficiency. Combining concurrency with inheritance is investigated in the framework of the actor computation model. A language design based on the actor model of concurrent computation faces a serious problem arising from the interference between concurrency and inheritance. A similar problem also occurs in other concurrent object-oriented languages. The problem of concurrency-inheritance conflict is described and a solution based on a concept called behavior abstraction is presented. / Ph. D.

Because software systems are imperfect, developers are forced to fix bugs and add new features. The common way of applying changes to a running system is to stop the application or machine and restart with the new version. Stopping and restarting causes a disruption in service that is at best inconvenient and at worst causes revenue loss and compromises safety. Dynamic software updating (DSU) addresses these problems by updating programs while they execute. Prior DSU systems for managed languages like Java and C# lack necessary functionality: they are inefficient and do not support updates that occur commonly in practice. This dissertation presents the design and implementation of Jvolve, a DSU system for Java. Jvolve's combination of flexibility, safety, and efficiency is a significant advance over prior approaches. Our key contribution is the extension and integration of existing Virtual Machine services with safe, flexible, and efficient dynamic updating functionality. Our approach is flexible enough to support a large class of updates, guarantees type-safety, and imposes no space or time overheads on steady-state execution. Jvolve supports many common updates. Users can add, delete, and change existing classes. Changes may add or remove fields and methods, replace existing ones, and change type signatures. Changes may occur at any level of the class hierarchy. To initialize new fields and update existing ones, Jvolve applies class and object transformer functions, the former for static fields and the latter for object instance fields. These features cover many updates seen in practice. Jvolve supports 20 of 22 updates to three open-source programs---Jetty web server, JavaEmailServer, and CrossFTP server---based on actual releases occurring over a one to two year period. This support is substantially more flexible than prior systems. Jvolve is safe. It relies on bytecode verification to statically type-check updated classes. To avoid dynamic type errors due to the timing of an update, Jvolve stops the executing threads at a DSU safe point and then applies the update. DSU safe points are a subset of VM safe points, where it is safe to perform garbage collection and thread scheduling. DSU safe points further restrict the methods that may be on each thread's stack, depending on the update. Restricted methods include updated methods for code consistency and safety, and user-specified methods for semantic safety. Jvolve installs return barriers and uses on-stack replacement to speed up reaching a safe point when necessary. While Jvolve does not guarantee that it will reach a DSU safe point, in our multithreaded benchmarks it almost always does. Jvolve includes a tool that automatically generates default object transformers which initialize new and changed fields to default values and retain values of unchanged fields in heap objects. If needed, programmers may customize the default transformers. Jvolve is the first dynamic updating system to extend the garbage collector to identify and transform all object instances of updated types. This dissertation introduces the concept of object-specific state transformers to repair application heap state for certain classes of bugs that corrupt part of the heap, and a novel methodology that employes dynamic analysis to automatically generate these transformers. Jvolve's eager object transformation design and implementation supports the widest class of updates to date. Finally, Jvolve is efficient. It imposes no overhead during steady-state execution. During an update, it imposes overheads to classloading and garbage collection. After an update, the adaptive compilation system will incrementally optimize the updated code in its usual fashion. Jvolve is the first full-featured dynamic updating system that imposes no steady-state overhead. In summary, Jvolve is the most-featured, most flexible, safest, and best-performing dynamic updating system for Java and marks a significant step towards practical support for dynamic updates in managed language virtual machines. / text

Abstraction mechanisms in programming languages aim to allow orthogonal pieces of functionality to be developed separately; complex software can then be constructed through the composition of these pieces. The effectiveness of such mechanisms lies in their support for modularity and reusability: The behavior of a piece of code should be reasoned about modularly---independently of the specific compositions it may participate in; the computation of a piece of code should allow specialization, so that it is reusable for different compositions. This dissertation introduces structural abstraction: a mechanism that advances the state of the art by allowing the writing of highly reusable code---code whose structure can be specialized per composition, while maintaining a high level of modularity. Structural abstraction provides a disciplined way for code to inspect the structure of its clients in composition, and declare its own structure accordingly. The hallmark feature of structural abstraction is that, despite its emphasis on greater reusability, it still allows modular type checking: A piece of structurally abstract code can be type-checked independently of its uses in compositions---an invaluable feature for highly reusable components that will be statically composed by other programmers. This dissertation introduces two structural abstraction techniques: static type conditions, and morphing. Static type conditions allow code to be conditionally declared based on subtyping constraints. A client of a piece of code can configure a desirable set of features by composing the code with types that satisfy the appropriate subtyping conditions. Morphing allows code to be iteratively declared, by statically reflecting over the structural members of code that it would be composed with. A morphing piece of code can mimic the structure of its clients in composition, or change its shape according to its clients in a pattern-based manner. Using either static type conditions or morphing, the structure of a piece of code is not statically determined, but can be automatically specialized by clients. Static type conditions and morphing both guarantee the modular type-safety of code: regardless of specific client configurations, code is guaranteed to be well-typed.

HOLCF is an interactive theorem proving system that uses the mathematics of domain theory to reason about programs written in functional programming languages. This thesis introduces HOLCF '11, a thoroughly revised and extended version of HOLCF that advances the state of the art in program verification: HOLCF '11 can reason about many program definitions that are beyond the scope of other formal proof tools, while providing a high degree of proof automation. The soundness of the system is ensured by adhering to a definitional approach: New constants and types are defined in terms of previous concepts, without introducing new axioms. Major features of HOLCF '11 include two high-level definition packages: the Fixrec package for defining recursive functions, and the Domain package for defining recursive datatypes. Each of these uses the domain-theoretic concept of least fixed points to translate user-supplied recursive specifications into safe low-level definitions. Together, these tools make it easy for users to translate a wide variety of functional programs into the formalism of HOLCF. Theorems generated by the tools also make it easy for users to reason about their programs, with a very high level of confidence in the soundness of the results. As a case study, we present a fully mechanized verification of a model of concurrency based on powerdomains. The formalization depends on many features unique to HOLCF '11, and is the first verification of such a model in a formal proof tool.

<p>Given the nature of client-server applications, most use some freeform interface, like web forms, to collect user input. The main difficulty with this approach is that all parameters obtained in this fashion need to be validated and normalized to protect the application from invalid entries. This is the problem addressed here: how to take client input and preprocess it before passing the data to a back-end, which concentrates on business logic. The method of implementation is a rule engine that uses Groovy internal domain-specific language (DSL) for specifying input requirements. We will justify why the DSL is a good fit for a validation rule engine, describe existing techniques used in this area and comprehensively address the related issues of accidental complexity, security, and user experience.</p> / Master of Science (MSc)

Modern data intensive applications, such as multimedia systems require the ability to store and manipulate complex data. The classical Database Management Systems (DBMS), such as relational databases, cannot support these types of applications efficiently. This dissertation presents the salient features of Object Database Management Systems (ODBMS) and Persistent Programming Languages (PPL), which have been developed to address the data management needs of these difficult applications. An 'impedance mismatch' problem occurs in the traditional DBMS because the data and computational aspects of the application are implemented using two different systems, that of query and programming language. PPL's provide facilities to cater for both persistent and transient data within the same language, hence avoiding the impedance mismatch problem. This dissertation presents a method of implementing a PPL by extending the language C++ with pre-compiled classes. The classes are first developed and then used to implement object persistence in two simple applications. / Computing / M. Sc. (Information Systems)