On 25 October 2012, at ACM’s SPLASH conference, six practitioners and academics came together for a panel discussion about “Software Tools Research A Matter of Scale and Scope—or Commoditization?” This episode is a postconference report on the discussions based a transcript of the session.

Listen:

A declarative programming style focuses on what you want your program to do rather than how to perform the task. By avoiding implementation details, well-written declarative code is easier to understand, modify, and maintain. The code you write in a declarative style is often so readable that you can share it with your project's domain experts. Furthermore, once you start working with declarative code, you can automatically process it to verify properties of its operation, generate test cases, or create parts of the system’s documentation. Small choices, such as the naming of your methods and variables, matter. If you’re coding an algorithm, have your code match the algorithm’s published description. Other approaches to consider include table-driven programming, the implementation of a domain-specific language, or the use of task-specific libraries. When you have the choice, pick the highest level language you can afford and that’s suitable for the task at hand.

Listen:

Open source software development efforts offer professionals a new and valuable way to obtain significant experience in a wide range of areas as an alternative to existing certification schemes.

Listen:

There's no language suitable for all tasks, and there probably won't ever be one. When choosing a programming language, you have to balance programmer productivity, maintainability, efficiency, portability, tool support, and software and hardware interfaces. Often, one of these factors will shape your decision. In other cases, the choice depends on the productivity you gain from certain language features, such as modularity and type checking, or external factors, such as integrated development environment support. Finally, for some tasks, adopting an existing domain-specific language, building a new one, or using a general-purpose declarative language can be the right choice.

Listen: