Artifacts developed alongside research presented in papers serve several important roles. They enable reproducing research results by independent researchers; they enable practically evaluating ideas and theories; they enable experimentation with variations and extensions (not only by the original authors); and so forth.
However, developing high-quality artifacts requires a lot of effort that is not always justified. Importantly, as it stands, artifacts have no influence on acceptance of a research paper at the top conferences in programming languages and software engineering. Moreover, the main requirement for completing a PhD is producing publishable research results. As a result, one of the most difficult questions for many doctoral students is: “how much time should be spent on developing artifacts?”
In this talk I share my experience contributing to the Scala programming language, tools, and libraries during my own graduate studies. Among the artifacts that I discuss: a library that was picked up by Twitter and secured a book deal; and a compiler plugin that (to my knowledge) only one other person in the world tried to use–and gave up. “How can artifacts support research and publication? How can artifacts support the discovery of important research problems? How can artifacts contribute to a successful career?” My talk attempts to shed light on these and related questions.
Assistant Professor in the School of Computer Science and Communication at KTH Royal Institute of Technology, Sweden.
Conference DayTue 27 OctDisplayed time zone: Eastern Time (US & Canada) change
15:30 - 17:20
|DS Invited Keynote Talk III: Papers vs. Artifacts|
Philipp HallerKTH Royal Institute of Technology
|Panini: A Concurrent Programming Model With Modular Reasoning |
|The Spreadsheet Paradigm: A Basis for Powerful and Accessible Programming|
A: Gary MillerUniversity of Technology Sydney