The growing popularity of aspect-oriented languages, such as AspectJ, and of corresponding design approaches, makes it important to learn how best to modularize programs in which aspect-oriented composition mechanisms are used. We contribute an approach to information hiding modularity in programs that use quantified advising as a module composition mechanism. Our approach rests on a new kind of interface: one that abstracts a crosscutting behavior, decouples the design of code that advises such a behavior from the design of the code to be advised, and that can stipulate behavioral contracts. Our interfaces establish design rules that govern how specific points in program execution are exposed through a given join point model and how conforming code on either side should behave. In a case study of the HyperCast overlay network middleware system, including a real options analysis, we compare the widely cited oblivious design approach with our own, showing significant weaknesses in the former and benefits in the latter.

Traditional design representations are inadequate for generalized reasoning about modularity in design and its technical and economic implications. We have developed an architectural modeling and analysis approach, and automated tool support, for improved reasoning in these terms. However, the complexity of constraint satisfaction limited the size of models that we could analyze. The contribution of this paper is a more scalable approach. We exploit the dominance relations in our models to guide a divide-andconquer algorithm, which we have implemented it in our Simon tool. We evaluate its performance in case studies. The approach reduced the time needed to analyze small but representative models from hours to seconds. This work appears to make our modeling and analysis approach practical for research on the evolvability and economic properties of software design architectures. 1.

According to Parnas’s information hiding principle and Baldwin and Clark’s design rule theory, the key step to decomposing a system into modules is to determine the design rules (or in Parnas’s terms, interfaces) that decouple otherwise coupled design decisions and to hide decisions that are likely to change in independent modules. Given a modular design, it is often difficult to determine whether and how its implementation realizes the designed modularity. Manually comparing code with abstract design is tedious and error-prone. We present an automated approach to check the conformance of implemented modularity to designed modularity, using design structure matrices as a uniform representation for both. Our experiments suggest that our approach has the potential to manifest the decoupling effects of design rules in code, and to detect modularity deviation caused by implementation faults. We also show that design and implementation models together provide a comprehensive view of modular structure that makes certain implicit dependencies within code explicit.

Modularity is one of the most important properties of a software design, with significant impact on changeability and evolvability. However, a formalized and automated approach is lacking to test and verify software design models against their modularity properties, in particular, their ability to accommodate potential changes. In this paper, we propose a novel framework for testing design modularity. The software artifact under test is a software design. A test input is a potential change to the design. The test output is a modularity vector, which precisely captures quantitative capability extents of the design for accommodating the test input (the potential change). Both the design and the test input are represented as formal computable models to enable automatic testing. The modularity vector integrates the net option value analysis with wellknown design principles. We have implemented the framework with tool supports and tested aspect-oriented and object-oriented design patterns in terms of their ability to accommodate sequences of possible changes. The results showed that previous informal, implementation-based analysis can be conducted by our framework automatically and quantitatively at the design level. This framework also opens the opportunities of applying testing techniques, such as coverage criteria, on software designs.

The Design Structure Matrix (DSM) methodology and the Net Option Value (NOV) model have been used before to show how aspects can add value to a design. Following with an in-depth analysis of that study, this paper demonstrates how aspects can be beneficial as well as detrimental. The structural transformations involved in aspect modularizations are carefully analyzed in the context of DSMs. This analysis exposes the unique reversion effect on dependencies that aspect modules are known for. To capture that effect within the NOV model, we extend its original set of six modular operators with an additional reversion operator. Using a design case study, its NOV worksheet and NOV experiments’ curves are presented to show a simulation of the evolutionary patterns of modules, including aspect modules. These patterns show how subtle dependencies, or the lack of them, bring down, or up, the value of an existing design. Based on the observations made in this case study, preliminary design guidelines for aspects are formulated.

There is some consensus in the aspect-oriented community that a notion of interface between joinpoints and advice may be necessary for improved modularity of aspect-oriented programs, for modular reasoning, and for overcoming pointcut fragility. Different approaches for adding such interfaces, such as aspect-aware interfaces, pointcut interfaces, crosscutting interfaces, explicit joinpoints, quantified typed events, open modules, and joinpoint types decouple aspects and base code, enhancing modularity. However, existing work has not shown how one can write specifications for such interfaces that will actually allow modular reasoning when aspects and base code evolve independently, and that are capable of specifying control effects, such as when advice does not proceed. The main contribution of this work is a specification technique that allows programmers to write modular specification of such interfaces and that allows one to understand such control effects. We show that such specifications allow typical interaction patterns, and interesting control effects to be understood and enforced. We illustrate our techniques via an extension of Ptolemy, but we also show that our ideas can be applied in a straightforward manner to other notions of joinpoint interfaces, e.g. the crosscutting interfaces.

Designers have long recognized the value of modularity, but important software modularity principles have remained informal. According to Baldwin and Clark’s [1] design rule theory (DRT), modular architectures add value to system designs by creating options to improve the system by substituting or experimenting on individual modules. In this paper, we examine the design evolution of two software product platforms through the modeling lens of DRT and design structure matrices (DSMs). We show that DSM models and DRT precisely explain how realworld modularization activities in one case allowed for different rates of evolution in different software modules and in another case conferred distinct strategic advantages on a firm by permitting substitution of an at-risk software module without substantial change to the rest of the system. Our results provide positive evidence that DSM and DRT can inform important aspects of large-scale software structure and evolution, having the potential to guide software architecture design activities. 1.

Modularity determines software quality in terms of evolveability, changeability, maintainability, etc. and a module could be a vertical slicing through source code directory structure or class boundary. Given a modularized design, we need to determine whether its implementation realizes the designed modularity. Manually comparing source code modular structure with abstracted design modular structure is tedious and error-prone. In this paper, we present an automated approach to check the conformance of source code modularity to the designed modularity. Our approach uses design structure matrices (DSMs) as a uniform representation; it uses existing tools to automatically derive DSMs from the source code and design, and uses a genetic algorithm to automatically cluster DSMs and check the conformance. We applied our approach to a small canonical software system as a proof of concept experiment. The results supported our hypothesis that it is possible to check the conformance between source code structure and design structure automatically, and this approach has the potential to be scaled for use in large software systems.

We were motivated to undertake the research we describe here by a conversation with two practicing software engineers, who described a dilemma they faced at work. They worked for small company that earned revenues by delivering

Abstract. We propose a formalization of crosscutting based on a conceptual framework for AOSD. Crosscutting is clearly distinguished from the related concepts scattering and tangling. The definitions of these concepts are formalized and visualized with matrices and matrix operations. This allows more precise reasoning about crosscutting.