www2.latech.eduLiving with the Lab Homepage - Tech

Watson Interactive Learning Suite eBook Labs Download About -- About History Acknowledgements History of Watson Watson development has taken place in four distinct phases: the first from 1993 through 1995, the second from 1996 to 1998, the third phase from 2002 to 2004, and the fourth phase from 2013 – 2014. Phase 1 (1993 – 1995) The first phase of Watson involved the conceptualization, design, and initial implementation of nine lab environments. During this time, Watson was support by a $100K grant from the National Science Foundation. A Watson Lab: Phase I (1993 – 1995) Watson’s core concepts of a unified graphical user interface, cross platform compatibility, breadth of concept coverage through the use of a suite of independent labs, and depth of coverage within individual labs by careful attention to the pedagogical content of individual lab activities, were established early on. The original suite of labs included: a simple spreadsheet, a relational database, a data structures lab, a graphics-based programming environment, an imperative programming environment, a functional programming environment, a Finite State Automata lab, an assembly language and machine architecture lab, and a digital logic lab. These labs were implemented in C and used the SUIT (Simple User Interface Toolkit) libraries for cross platform compatibility. While these prototypes established the viability of the Watson approach to introductory computer science education, they were not robust enough to be used outside of a closed laboratory environment with extensive student assistance provided by instructors and lab monitors. In fact, the Project Director’s candid assessment of the original implementation of Watson would be “buggy as hell”. This should not be surprising, given the goals of the project, its groundbreaking nature, and limited funding. Additionally the interface provided by the labs were not very attractive, nor intuitive. Phase II (1996-1998) By the mid-1990’s it became apparent that Watson needed a total rewrite in order to become a viable teaching tool. After several years of continuous classroom use, resulting in feedback from hundreds of students, the Project Director had a very clear idea of what worked and what didn’t. Furthermore, the emergence of Java and explosion of web technologies offered the hope of a stable cross-platform environment for Watson. Given that NSF funding had expired at this point, the Project Director, set about the bold task of designing and implementing Watson in Java. Since very few Java-based tools or libraries existed at the time (combined with the fact that Java itself was rapidly evolving), the Project Director began by building up a “Watson Java Toolkit” of common components, such as choice selectors, numeric entry pads, and even low-level primitives such as “Watson Buttons”. Once these were in place he redesigned the graphics lab based on student feedback from the Phase I version. He then coded the lab and thoroughly debugged it. The graphics lab thus served as a “template” for all Watson Phase II labs. Given this template students were offered independent study classes in which they would “re-implement” a Phase I lab using the toolkit and graphics lab as a guide. Some of the labs such as spreadsheet, database, assembly, and digital logic incorporated few pedagogical changes in their port to Java. Other labs, such as data structures, graphics, and imperative programming were completely rethought. Still other labs, such as the functional programming lab and finite state automata lab, were abandoned. A Watson Lab: Phase II (1996 – 1998) The result of Phase II was a collection of stable, albeit somewhat plain-looking labs, that could be accessed via web browsers. These labs were used successfully in Louisiana Tech’s CSC 100 courses for a number of years. By 1998, the project had reached the point where national dissemination appeared imminent. In addition to Watson, its companion textbook (drafts of which have been student tested since the mid-1990’s) was nearing completion. Despite Watson’s promise, the Project Director had to reluctantly shelve Watson for several years. The continued lack of external funding for Watson, combined with allure of the dot com boom, enticed the Project Director into a two-year sabbatical in industry where I became the Chief Technology Officer for OneNetNow.com. After OneNetNow was sold to Earthlink in 2001, the Project Director elected to return to academia rather than accept a permanent position with EarthLink. Phase III (2002-2004) Upon return to academia in the Fall of 2001, four things became apparent (1) Since the labs had been written in the very early days of Java, they were no longer stable on modern Java environments, (2) The look of the labs had become outdated since they were not built with Swing components, (3) The pedagogical content embodied in the Imperative lab was no longer as appropriate as it had been in the mid-1990’s, and (4) The fact that Watson contained zero coverage of object-oriented programming concepts had become a serious omission. Additionally, the uncompleted textbook could not simply be finished, as sections of it needed to be rewritten to reflect the changes in the field from the mid-late 1990’s to the early 2000’s. Furthermore, since the textbook and the labs are interrelated, any changes to the labs required similar changes to the text. The Project Director began restarting Watson development in 2002. He recruited a graduate student, Cliff Lemoine, to help maintain the Phase II labs and to design and develop an object-oriented programming lab. In the Summer of 2002, he made some inroads into updating the textbook. During the Spring of 2003, he applied to the National Science foundation for funding to fully launch Phase III. Though the proposal reviews were generally strong, the NSF ultimately declined the application. (Oh well… Watson is not so easily killed.) So, in the tradition of Watson, the Project Director embarked on Phase III. He recruited a bright team of undergraduates and with nothing but Independent Study credit in his pocket and pushed forward with Phase III. The Phase III team was able to build on a Swing-based Watson platform implemented in 1999 by a former graduate student, Don Garrett, together with a nearly complete implementation of the graphics lab Don wrote. The Phase III team made rapid progress re-implementing several of the labs. From December 2003 through February 2004, database, assembly, and digital logic were ported. Data structures, and JavaScript were begun. From March 2004 to May 2004, the spreadsheet lab was successfully ported. During this time, the data structures rewrite was completed, and the initial implementation of the object-oriented lab was finished. As of May 2004 the only lab that is incomplete is the JavaScript lab. Unfortunately, from a pedagogical standpoint, this is one of the most important Watson labs. A Watson Lab: Phase III (2003-2004) During the spring of 2004, the Project Director had the good fortune of teaching CSC 404 “Senior Capstone”. In the spirit of “use every resource at your disposal”, the Project Director selected Watson as the topic for Senior Capstone. In addition to providing desperately needed manpower for the Watson project, this choice enabled the class to gain near-real-world experience with a large project on the back end of the software design process. The CSC 404 class devoted tremendous effort to testing, documentation, and code maintenance. They tested the Phase III software with actual students and incorporated UI changes based on the student’s feedback. Furthermore the class refined the graphical “look” of the labs and incorporated this look into the web and printed documentation they developed. As of May 2004, the Project Director was once again in the process of applying to the National Science Foundation for funds. These...