David G. Wonnacott

• CS105 Introduction to Computer Science
  Lecture MW 12:30-2:00 in Hilles 109; Lab any one hour of Th 7-8PM, Th 8-9PM, F 12-1PM, F 1-2PM in Hilles 110 (possibly with one more section to be organized).
• CS240 Principles of Computer Organization
  Lecture T/Th 1:00-2:30 in KINSC E309; Lab F 2:30-4:00 in Hilles 110
• CS460 Teaching Assistantship in Computer Science (with J.D.)
  Group meetings Wednesday 2:30-3:30.
• Senior research/reading group: Friday 10-11AM
• Office/Lab Hours (in KINSC H110 unless otherwise noted):
  • M 2:00-2:30 or 3:00PM (depending on attendance)
  • M 8:00-9:00 PM
  • Tu 11:00AM-noon
  • W 2:00-2:30 PM
  • Th 2:30-3:30 PM in KINSC L306, with priority for CMSC 240 students
• Spring 2009: On leave

Current Research

• Computer Science Education
  • J.D. and I have been working on revising the introductory year of computer science to emphasize the intellectual tools computer scientists use to analyze algorithms and data structures. This was a continuation of the direction that had already been taken by Steven Lindell and the others involved in setting up computer science at Haverford.

    The goals of our work are to better prepare students who wish to continue in computer science, provide a better appreciation of the intellectual processes involved in CS for students who do not continue, and emphasize the fun of problem solving over the tedium of learning the arcane details of a programming language. After exploring the use of mathematical techniques (as used in formal verification), we have more recently (post-2005) been working on a balance between informal and formal approaches, and between test suites and specifications.

    For more details, see the the 3-2-1 web site, or our paper at SIGCSE '05 (for our work before 2005) or the current course notes at lulu.com. (We have been quite happy with the quality and price of lulu.com materials: the "lab copy" of the materials that we put in the lab in September 2007 was still in good condition when the academic year ended in May 2008).

  • I have developed HERA, the Haverford Educational RISC Architecture, to provide a single microprocessor specification that is powerful enough to be used to learn assembly language programming in Principles of Programming Languages and to serve as a target for the compilers written in Compiler Design, while still being simple enough to be built by students in Principles of Computer Organization and Operating Systems.

    The goal of this work is to help students appreciate the connections among these upper-level applied computer science courses by constructing a single multi-course project that involves all the steps from a high-level programming language to the basic AND/OR/NOT logic of computer hardware. The first students to complete this entire sequence using HERA graduated in the spring of 2005.

    I presented this work in a poster session (not reviewed) at SIGCSE '05. A (reviewed) paper describing this work and another describing my use of attribute grammars in teaching Compiler Design appeared at CCSCE '06. The specification for the HERA architecture is available in printed form through lulu.com.

• Visualization of algorithm execution.

  I have recently begun to investigate techniques to help visualize the execution of algorithms -- if my own ideas prove to be sufficiently different from prior work on algorithm animation, I'll add more here.

• Uses of compiler technology to inform or direct system services.

  I have worked with the Vivo project at the Rutgers University Department of Computer Science to develop tools to instrument code to measure and improve coverage of fault recovery code in Java web services (see our papers at DSN '03, ISSTA '04, and The April 2005 issue of IEEE Transactions on Software Engineering).

• Constraint-based analysis of calculations involving arrays (the Omega Test).

  Vince Gabor (HC '99) and Rob Seater (HC '02) participated in this work.

• Optimizing regular array calculations for memory locality (Time Skewing and Scalable Locality).

  This began as a join investigation with John McCalpin, and involved numerous undergraduates, including Jaime Spacco (HC '98), Tina Shen (Bryn Mawr '98), Vince Gabor (HC '99), and Frank Rusch (HC 2000). Recently, Ian Harrison (HC '03) and I have gotten some very nice empirical results for the multiprocessor time skewing optimization on an 8-workstation linux cluster.

• Adapting scalar optimizations for array variables .

• Reconciling encapsulation and dispatch mechanisms in object-oriented languages that support data encapsulation (Accessory Functions).

  This is an extension of Ben Flynn's (HC '99) senior thesis, which appeared at MASPLAS '99.

The above research has been funded by grants from the Zimmer Corporation and grant CCR-9808694 from the National Science Foundation. "Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation."

Additional Information:

Here's a list of Researchers in Programming Languages and Compilers.

I use Gnu Privacy Guard here's my public key.