Classless Object Semantics

<p>Objects have been categorised into classes that declare and implement their behaviour ever since the paradigm of object-orientation in programming languages was first conceived. Classes have an integral role in the design and theory of object-oriented languages, and often appear alongside objects as a foundational concept of the paradigm in many theoretical models. A number of object-oriented languages have attempted to remove classes as a core component of the language design and rebuild their functionality purely in terms of objects, to varying success. Much of the formal theory of objects that eschews classes as a fundamental construct has difficulty encoding the variety of behaviours possible in programs from class-based languages. This dissertation investigates the foundational nature of the class in the object-oriented paradigm from the perspective of an ‘objects-first’, classless language. Using the design of theoretical models and practical implementations of these designs as extensions of the Grace programming language, we demonstrate how objects can be used to emulate the functionality of classes, and the necessary trade-offs of this approach. We present Graceless, our theory of objects without classes, and use this language to explore what class functionality is difficult to encode using only objects. We consider the role of classes in the types and static analysis of object-oriented languages, and present both a practical design of brand objects and a corresponding extension of our theory that simulates the discipline of nominal typing. We also modify our theory to investigate the semantics of many different kinds of implementation reuse in the form of inheritance between both objects and classes, and compare the consequences of these different approaches.</p>

Classless Object Semantics

<p>Objects have been categorised into classes that declare and implement their behaviour ever since the paradigm of object-orientation in programming languages was first conceived. Classes have an integral role in the design and theory of object-oriented languages, and often appear alongside objects as a foundational concept of the paradigm in many theoretical models. A number of object-oriented languages have attempted to remove classes as a core component of the language design and rebuild their functionality purely in terms of objects, to varying success. Much of the formal theory of objects that eschews classes as a fundamental construct has difficulty encoding the variety of behaviours possible in programs from class-based languages. This dissertation investigates the foundational nature of the class in the object-oriented paradigm from the perspective of an ‘objects-first’, classless language. Using the design of theoretical models and practical implementations of these designs as extensions of the Grace programming language, we demonstrate how objects can be used to emulate the functionality of classes, and the necessary trade-offs of this approach. We present Graceless, our theory of objects without classes, and use this language to explore what class functionality is difficult to encode using only objects. We consider the role of classes in the types and static analysis of object-oriented languages, and present both a practical design of brand objects and a corresponding extension of our theory that simulates the discipline of nominal typing. We also modify our theory to investigate the semantics of many different kinds of implementation reuse in the form of inheritance between both objects and classes, and compare the consequences of these different approaches.</p>

Multiple Dispatch in Practice

<p>Multiple dispatch uses the run time types of more than one argument to a method call to determine which method body to run. While several languages over the last 20 years have provided multiple dispatch, most object-oriented languages still support only single dispatch - forcing programmers to implement multiple dispatch manually when required. This thesis presents an empirical study of the use of multiple dispatch in practice, considering six languages that support multiple dispatch. We hope that this study will help programmers understand the uses and abuses of multiple dispatch; virtual machine implementors optimise multiple dispatch; and language designers to evaluate the choice of providing multiple dispatch in new programming languages.</p>

Multiple Dispatch in Practice

<p>Multiple dispatch uses the run time types of more than one argument to a method call to determine which method body to run. While several languages over the last 20 years have provided multiple dispatch, most object-oriented languages still support only single dispatch - forcing programmers to implement multiple dispatch manually when required. This thesis presents an empirical study of the use of multiple dispatch in practice, considering six languages that support multiple dispatch. We hope that this study will help programmers understand the uses and abuses of multiple dispatch; virtual machine implementors optimise multiple dispatch; and language designers to evaluate the choice of providing multiple dispatch in new programming languages.</p>

Evaluation of Rust code verbosity, understandability and complexity

Rust is an innovative programming language initially implemented by Mozilla, developed to ensure high performance, reliability, and productivity. The final purpose of this study consists of applying a set of common static software metrics to programs written in Rust to assess the verbosity, understandability, organization, complexity, and maintainability of the language. To that extent, nine different implementations of algorithms available in different languages were selected. We computed a set of metrics for Rust, comparing them with the ones obtained from C and a set of object-oriented languages: C++, Python, JavaScript, TypeScript. To parse the software artifacts and compute the metrics, it was leveraged a tool called rust-code-analysis that was extended with a software module, written in Python, with the aim of uniforming and comparing the results. The Rust code had an average verbosity in terms of the raw size of the code. It exposed the most structured source organization in terms of the number of methods. Rust code had a better Cyclomatic Complexity, Halstead Metrics, and Maintainability Indexes than C and C++ but performed worse than the other considered object-oriented languages. Lastly, the Rust code exhibited the lowest COGNITIVE complexity of all languages. The collected measures prove that the Rust language has average complexity and maintainability compared to a set of popular languages. It is more easily maintainable and less complex than the C and C++ languages, which can be considered syntactically similar. These results, paired with the memory safety and safe concurrency characteristics of the language, can encourage wider adoption of the language of Rust in substitution of the C language in both the open-source and industrial environments.

On theories of names and references in formal languages and implications for functional and object-oriented programming

The long-standing problem of adequate formalization of local names in mathematical formulae and semantics of references in object-oriented languages taken “as is” without objects, is discussed. Reasons why the existing approaches cannot be considered suitable solutions, are explained. An introduction to the relatively recent works on the theories of names and references of the group headed by Andrew Pitts, is given. The notion of referential transparency, in which contextual equivalence is used instead of the usual equality of values, is analyzed. This is the main property, which these theories are based upon: it is preserved when a purely functional language is extended with names and references as data. It is argued that such referential transparency, along with many others, can be preserved for mutable objects that change to a limited extent. This leads to a model of computation between functional and object-oriented ones, allowing for a deterministic parallel implementation.

BioTracs: A transversal framework for computational workflow standardization and traceability

Background: The need of digital tools for integrative analysis is today important in most scientific areas. It leads to several community-driven initiatives to standardize the sharing of data and computational workflows. However, there exists no open agnostic framework to model and implement computation workflows, in particular in bioinformatics. It is therefore difficult for data scientists to share transparently and integrate heterogeneous analysis processes coming from different scientific domains, programing languages, projects or teams. Results: We present here BioTracs, a transversal framework for computational workflow standardization and traceability. It is based on PRISM architecture (Process Resource Interfacing SysteM), an agnostic open architecture we introduce here to standardize the way processes and resources can be modelled and interfaced in computational workflows to ensure traceability, reproducibility and facilitate sharing. BioTracs is today implemented in MATLAB and available under open source license on GitHub. Several BioTracs-derived applications are also available online. They were successfully applied to large-scale metabolomics and clinical studies and demonstrated flexibility and robustness. Conclusions: As an implementation of the PRISM architecture, BioTracs paved the way to an open framework in which bioinformatics could specify ad model workflows. PRISM architecture is designed to provide scalability and transparency from the code to the project level we less efforts. It could also be implemented using open object-oriented languages such as Python, C++ or java.

BioTracs: A transversal framework for computational workflow standardization and traceability

Background The need of digital tools for integrative analysis is today important in most scientific areas. It leads to several community-driven initiatives to standardize the sharing of data and computational workflows. However, there exists no open agnostic framework to model and implement computation workflows, in particular in bioinformatics. It is therefore difficult for data scientists to share transparently and integrate heterogeneous analysis processes coming from different scientific domains, programing languages, projects or teams.Results We present here BioTracs, a transversal framework for computational workflow standardization and traceability. It is based on PRISM architecture (Process Resource Interfacing SysteM), an agnostic open architecture we introduce here to standardize the way processes and resources can be modelled and interfaced in computational workflows to ensure traceability, reproducibility and facilitate sharing. BioTracs is today implemented in MATLAB and available under open source license on GitHub. Several BioTracs-derived applications are also available online. They were successfully applied to large-scale metabolomics and clinical studies and demonstrated flexibility and robustness.Conclusions As an implementation of the PRISM architecture, BioTracs paved the way to an open framework in which bioinformatics could specify ad model workflows. PRISM architecture is designed to provide scalability and transparency from the code to the project level we less efforts. It could also be implemented using open object-oriented languages such as Python, C++ or java.

Chatty Ship: Disseminated, Entrenched Information

Stable algorithms and e-commerce have gar-nered limited interest from both cryptogra-phers and electrical engineers in the last sev-eral years. Though such a hypothesis is usu-ally a robust purpose, it is buffetted by prior work in the field. After years of confusing re-search into RAID, we prove the refinement of object-oriented languages, which embodies the natural principles of topologically extremely stochastic e-voting technology. In this position paper, we propose an analysis of public-private key pairs (ChattyShip), showing that active net-works [10, 14] and wide-area networks can in-teract to fulfill this purpose.

A Case for I/O Automata

Physicists agree that empathic archetypes is a new topic in the area of network-ing, and physicists concur. After years of ro-bust research into object-oriented languages, we demonstrate the development of extreme pro-gramming, which embodies the extensive princi-ples of complexity theory. Our focus here is not on whether neural networks and fiber-optic ca-bles can interact to address this quagmire, but rather on exploring a robust tool for enabling 802.11b (Marten).