Object-oriented frameworks are a popular mechanism for building and evolving large applications and software product lines. This paper describes an alternative approach to software construction, Java Layers (JL), and evaluates JL and frameworks in terms of flexibility, ease of use, and support for evolution. Our experiment compares Schmidt's ACE framework against a set of ACE design patterns that have been implemented in JL. We show how problems of framework evolution and overfeaturing can be avoided using JL's component model, and we demonstrate that JL scales better than frameworks as the number of possible application features increases. Finally, we describe how constrained parametric polymorphism and a small number of language features can support JL's model of loosely coupled components and stepwise program refinement.

Comparing frameworks and layered refinement

Testing is among the most effective tools available for finding bugs. Still, we know of no automatic technique for generating test cases that expose bugs involving a combination of mutable state and callbacks, even though objects and method overriding set up exactly that combination. For such cases, a test generator must create callbacks or subclasses that aggressively exercise side-effecting operations using combinations of generated objects.This paper presents a new algorithm for randomly testing programs that use state and callbacks. Our algorithm exploits a combination of contracts and environment bindings to guide the test-case generator toward interesting inputs. Our prototype implementation for Racket (formerly PLT Scheme) - which has a Java-like class system, but with first-class classes as well as gbeta-like augmentable methods - uncovered dozens of bugs in a well-tested and widely used text-editor library.We describe our approach in a precise, formal notation, borrowing the techniques used to describe operational semantics and type systems. The formalism enables us to provide a compact and self-contained explanation of the core of our technique without the ambiguity usually present in pseudo-code descriptions.

/pdf/random-testing-for-higher-order-stateful-programs-3mpuoc5s1e.pdf

Random testing for higher-order, stateful programs

The emergence of e-marketplace Web sites that contain proprietary information from multiple organizations requires the creation of new access control schemes that provide fine-grained access control while reducing both administrative and run-time overhead. It is also desirable to have clear, concise, and easily configurable definitions of access control policies that are aligned with business processes, and to have these policies enforced consistently throughout an ecommerce system. In this paper, we describe a policy-based access control scheme, and its implementation, that allows access to individual instances of resources to be specified in a concise and computationally efficient manner. We model business relationships between users and business objects and use implicit grouping of users and resources. These concepts allow policies to refer efficiently to aggregates of resources and users and to document the intention of an authorization policy. Our access control scheme is implemented as an application-level access control mechanism within IBM's WebSphere® Commerce Suite, Marketplace Edition. We use this implementation to provide examples and give performance data. For future work, we discuss how our policy-based, resource-level access control scheme might be enhanced to augment language-level access control schemes, such as the Java™ 2 Platform, Enterprise Edition (J2EE™) security model.

Instance-level access control for business-to-business electronic commerce

The invention provides a system and method for constructing and maintaining a price structure for an e-commerce transaction relating to a complex product or service. A pricing tree is built of “price components” which include discrete revenues, price adjustments, taxes, and disbursements (such as royalties). Separate revenue streams, which may be allocated to different service providers, may have independent adjustment, tax, and disbursement treatments. Attributes may also be used in the calculation of separate components. Time-date adjustments and rate package adjustments are also supported. The system allows both automated calculation and accounting using the pricing tree. Transactions are supported by a modular e-commerce back-end platform.

Method and system for constructing price structures for complex products and services

One of the obstacles preventing the widespread adoption of multi-agent systems in industry is the difficulty of implementing heterogeneous interactions among participating agents via asynchronous messages. This difficulty arises from the need to understand how to combine elements of various content languages, ontologies, and interaction protocols in order to construct meaningful and appropriate messages. In this paper mPower, a component-based layered framework for easing the development of multi-agent systems, is described, and the facility for customising the components for reuse in similar domains is explained. The framework builds on the JADE-LEAP platform, which provides a homogeneous layer over diverse operating systems and hardware devices, and allows ubiquitous deployment of applications built on multi-agent systems both in wired and wireless environments. The use of the framework to develop mPowermobile, a multi-agent system to support mobile workforces, is also reported.

/pdf/mpower-a-component-based-development-framework-for-multi-4kz6bclh87.pdf

mPower -- a component-based development framework for multi-agent systems to support business processes

The San FranciscoTM project establishes a new paradigm for building business applications. The product, targeted for independent software vendors (ISVs), provides a distributed object infrastructure (foundation), common business objects (CBOs), and business process components (BPCs). Together, they provide a platform-independent business application foundation, ready for extension by ISVs to produce end-customer, business-critical applications. The San Francisco project is written almost entirely in JavaTM and to our knowledge is currently the largest Java development effort in the world. This paper provides an overview of the San Francisco project, with emphasis on the Java considerations of the product's development, the lessons learned, and our recommendations for future Java language maturity.

A. R. Christ

Papers

Java and the IBM San Francisco project