Panagiotis Louridas

Power laws in software

A single statistical framework, comprising power law distributions and scale-free networks, seems to fit a wide variety of phenomena. There is evidence that power laws appear in software at the class and function level. We show that distributions with long, fat tails in software are much more pervasive than previously established, appearing at various levels of abstraction, in diverse systems and languages. The implications of this phenomenon cover various aspects of software engineering research and practice.

Using wikis in software development

Wikis are especially useful in distributed projects: many teams around the world use them to organize, track, and publish their work. Their flexibility frees a project manager from fretting about getting everything exactly right from the beginning. A wiki can and should change to respond to the projects needs as they arise. While wikis are always easy to change, wiki engines usually incorporate comprehensive versioning and change control for their content. More important than the particular wiki implementation, however, is being sure that a wiki really fits in the culture of the project or organization. It requires tolerance and openness.

The collaborative organization of knowledge

Why Wikipedias remarkable growth is sustainable.

Up in the Air: Moving Your Applications to the Cloud

An overview of cloud computing helps developers get beyond the hype by characterizing its unique requirements, the ways to implement cloud services, and the ways to operate the cloud infrastructure. A comparison of major products is included.

Evaluating the Quality of Open Source Software

Traditionally, research on quality attributes was either kept under wraps within the organization that performed it, or carried out by outsiders using narrow, black-box techniques. The emergence of open source software has changed this picture allowing us to evaluate both software products and the processes that yield them. Thus, the software source code and the associated data stored in the version control system, the bug tracking databases, the mailing lists, and the wikis allow us to evaluate quality in a transparent way. Even better, the large number of (often competing) open source projects makes it possible to contrast the quality of comparable systems serving the same domain. Furthermore, by combining historical source code snapshots with significant events, such as bug discoveries and fixes, we can further dig into the causes and effects of problems. Here we present motivating examples, tools, and techniques that can be used to evaluate the quality of open source (and by extension also proprietary) software.

Static code analysis

Programmers usually employ static checkers, it checks our programs for errors without executing them, in a process called static code analysis. In this way, it works with a program that has an initial indication of correctness (because it compiles) and try to avoid well-known traps and pitfalls before measuring it against its specifications (when its tested). We use FindBugs, a popular open source static code checker for Java. Static code checkers in Java come in two flavors: those that work directly on the program source code and those that work on the compiled bytecode. Although each code checker works in its own way, most share some basic traits. They read the program and construct some model of it, a kind of abstract representation that they can use for matching the error patterns they recognize. They also perform some kind of data-flow analysis, trying to infer the possible values that variables might have at certain points in the program. Data-flow analysis is especially important for vulnerability checking, an increasingly important area for code checkers

Orchestrating Web Services with BPEL

Effective Web services demand careful synchronization on various abstraction levels. The Business Process Execution Language supports modeling and executing business processes from both the user and systems perspectives. In this way, Web services application developers can use BPEL to orchestrate service interactions in a global system view and to manage individual interactions based on outside events. More and more Web service providers are using BPEL to integrate their services independently of vendors and related programming languages. In this column, we introduce BPEL and presents some guidelines for using it in your Web services management.

Some guidelines for non-repudiation protocols

Non-repudiation protocols aim at preventing parties in a communication from falsely denying having taken part in that communication; for example, a non-repudiation protocol for digital certified mail should ensure that neither the sender can deny sending the message, nor the receiver can deny receiving it. We identify some guidelines for non-repudiation protocols. The guidelines are derived by examining a series of non-repudiation protocols that descend from a single ancestor.

JUnit: unit testing and coiling in tandem

Detecting and correcting defects either during or close to the phase where they originate is key for any fast, cost-effective software development. Unit testing, performed by the code writer, is still the most popular technique. The author has summarized the experiences around JUnit, probably the most popular OSS tool for Java unit testing. JUnit is an open source Java library, Kent Beck and Erich Gamma created JUnit, deriving it from Becks Smalltalk testing framework. Easy integration of coding and unit testing lies at the heart of Extreme Programming and the agile software development movement. In agile programming, software is constructed by implementing functionality incrementally, in short spurts of activity.

Version control

Software evolves in small steps or versions. Often these versions are results of collaborations among different persons. At times we want to fall back to a previous version or compare different variants. Or, we need to trace changes to change requests. All these tasks relate to version control and show that tool support in this domain is indispensable, as it is with editors and compilers. The author describes available tools and shows why the CVS open source tool is so popular among practitioners