blog dds

2013.03.14

Software Tools Research: SPLASH Panel Discussion

Written by Dennis Mancl and Steven Fraser

At the recent SPLASH (Systems, Programming, Languages and Applications: Software for Humanity) conference, one of us (Steven Fraser) organized an international group of experts to discuss challenges in software tools research.1 The panelists included Kendra Cooper (University of Texas, Dallas), Jim “Cope” Coplien (Gertrud & Cope), Junilu Lacar (Cisco Systems), Ruth Lennon (Letterkenny Institute of Technology), Diomidis Spinellis (Athens University of Economics and Business), and Giancarlo Succi (Free University of Bolzano-Bozen).

The discussion interwove three threads—tool use, development, and education—and the panelists took a critical look at how well tools serve the needs of software professionals, managers, and academics. Their passion for the topic was reflected through some heated exchanges, even during the opening statements.

Tool Use and Misuse

Cooper contrasted the perspective of tool users in industry who are building and integrating large production systems with the researchers who build small prototype tools to explore software engineering ideas.

Lacar spoke for industry: he wants tools research to focus more on creating tools and languages that are clean, clear, and simple. Tools must be focused on the job tasks of software professionals; they must help individual developers think and support clear communication within development teams.

Succi described research that collects data about how tools are really used. Most software developers only use eight to 12 tools, even when more tools are available. Succi believes that tool users should reflect on their tool usage and workflows, and tools should report on how they’re used. This would allow developers to assess the real “business value” of their tools and give them input on how to adjust the tools and their development processes to produce value more effectively.

Cope complained about the misuse of tools: “The way managers manage people is through adherence to the tools. So it isn’t that the tools guide the way, they enforce the managers’ will to implement workflows conceived 10 years ahead of the time that they’re used, that are 10 years out of date with prevailing knowledge and maturity. Tools are the enforcer of the management prisons, they’re the guards at the gates of the prisons, to keep the developers from being innovative.”

Spinellis concurred, lamenting the fact that his UMLGraph tool, which allows designers to create simple UML diagrams from a simple text-based specification language, was hijacked by its user community to generate UML diagrams directly from Java code.

From there, the conversation heated up. Cope criticized the software development community for relying on tools, especially tools to support the design process. Commodity tools are fine, such as simple program editors and compilers for building software: they’re the equivalent of hammers and saws for building houses. But when inexperienced software designers use design tools based on UML, they can easily confuse a pretty picture with a good quality design. When managers mandate agile management tools such as RallyDev, they might use the tool to enforce a non-agile waterfall development process. Cope went on to declare that tools are evil because they encourage misuse: software professionals doing their jobs without thinking, learning, and communicating with others. “We don’t need more research on tools, we need to get rid of the tools,” he stated.

Lennon immediately pointed out the inconsistency of Cope’s complaints about tool use: “While you’re giving out about tools, you’re looking at your laptop and using a tool.”

Succi also challenged Cope’s statement that tools are always bad by pointing out that the existing research literature on how users really use their tools is very sparse. Cope brought up some of his work on using games to get team members to communicate and collaborate more effectively—for example, the ScrumKnowsy tool, a game designed to encourage more collaboration within a Scrum team.

Education

Lennon spoke for the learning community: the students at universities and technical institutes who use tools. The community of learners is large and diverse. Younger learners are already comfortable with using tools such as BitTorrent, YouTube, and Dropbox on a daily basis. More institutions are using virtual learning environments to allow learners to more conveniently access study materials. Some tools support hands-on learning experiences, whereas others permit students to view or listen to short 10-minute learning nuggets on the learner’s own schedule.

Succi confessed his preference to do serious software development with minimal tools, while teaching development with more sophisticated tools: “When I have anything serious to do, I use vi (Unix text editor). However, when I have to train a young person in class, I use Eclipse.”

Questions from the floor raised more issues about learning and communication. There’s a lot of interest in simple Web-based learning tools, following the lead of Minute Physics and Khan Academy. Cooper discussed the “gamification” of applications: “I’ve been working on games for a while, serious educational games, and I see that they’re trying to engage people and get people more involved with what they’re working with.”

Tool Development

Spinellis shared some of his experiences as a developer of open source software development tools. The increasing complexity of software systems, programming languages, and development environments is making the development of new tools more difficult. On the other hand, development and delivery environments like github support the sharing of open source tools and incorporate incremental improvements made by tool users. The best tools aren’t big monoliths, they’re small bricks that others can use to build on.

David Ungar, speaking from the floor, suggested that most software tools violate the object-oriented design principle that behavior should be combined with data. The Self language, developed by Ungar and Randall B. Smith at Xerox PARC in the late 1980s, used a different philosophy: instead of having an inspector to look at an object, Self just had the object appear on the screen as an outliner. So functionality that would normally be in a separate tool was part of the environment: “Every bolt had a handle on it, instead of needing a wrench.” Ungar’s approach can’t replace all tools, but during the discussion, Spinellis pointed out that it’s good to build more tools that do one thing well and can be composed with other tools.

Lacar addressed a question about “forking,” where multiple versions of an open source tool are created in parallel. Forking is mostly good because it leads to competition and improvement, but it also causes a lot of confusion. When there’s a fork in a popular tool, such as the Hudson/Jenkins continuous integration build management systems, users are left wondering, “What do we do now? Which way do we go?”

Lennon’s final comment summarized the view of the panel’s academics. To do exploratory work on tools, researchers need to go beyond a handful of small case studies, but companies need to be more open to sharing information: “If you don’t trust, we can’t get in and do the research. If you don’t trust us to keep your data confidential, we’ll never get there.”

A complete transcript of the panel discussion can be found through this link.

Reference

1. S. Fraser et al., “Software Tools Research: A Matter of Scale and Scope—or Commoditization?” Proc. Conf. Systems, Programming, and Applications: Software for Humanity (SPLASH 12), G.T. Leavens, ed., ACM, 2012, pp. 59–62; doi:10.1145/2384716.2384740.

* This piece has been published in the IEEE Software magazine Tools of the Trade column, and should be cited as follows: Dennis Mancl and Steven Fraser Software Tools Research: SPLASH Panel Discussion. IEEE Software, 30(2):16–17, March/April 2013. (doi:10.1109/MS.2013.36)

Read and post comments, or share through   


© 2013 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Last modified: Sunday, June 16, 2013 2:13 pm