An Architecture Promoting Collaborative Research, Teaching, and Learning

The organization of a complex system like the Edition Production Technology (EPT) presents a set of serious
design problems. Its workbench must offer comprehensive services, which the editor and co-researchers need
to prepare and maintain DTDs or X-Schemas, encode texts with searchable descriptive and editorial markup
in XML, integrate images, and design interfaces. It must also prepare powerful interactive displays for people
to make full use of the completed image-based edition. Unlike most software projects, moreover, whose
primary objective is to produce a functional and robust application, ARCHway’s goals are more complicated,
requiring an infrastructure that integrates Research, Teaching and Learning between the very different
disciplines of computer science and the humanities. To meet this principal purpose, we must carefully plan
the underlying architecture of EPT, which will guide the development process and map the functionality into
the overall system. Our architectural model promotes the development of modular, extensible, interoperable
components in a collaborative research and teaching environment that actively involves researchers and
students in both disciplines.
RESEARCH
EPT is an innovative system whose strengths hinge on a number of ideas that require novel solutions and
utilization of emerging technologies. Among them are questions of effective, convenient, and complete
mapping of the phenomena of the editorial domain into a software system. They lead to research problems of
image archiving, image access and manipulation, methods for automated linking of images with the text,
storage, access, secure transactions, data persistence and consistency (algorithms), and virtual remorphing.
The intellectual integrity of the project will be achieved by continuous exchange of “what” and “how”
knowledge and domain specific-phenomena between humanities scholars and computer and technology
specialists. The workbench that we create will mirror and augment the processes, methodologies, and
software tools the editor uses to develop electronic editions from original manuscripts.
Reminiscent of the currently prevailing Windows manager interface introduced by Doug Englebart
and his Stanford team in the late sixties, the workbench will attempt to recreate a virtual environment for
studying codicology. As Frederick Brooks, an authority in software engineering puts it, “The Windows,
Icons, Menus and Pointing interface is a superb example of a user interface that has conceptual integrity,
achieved by the adoption of a familiar mental model, the desktop metaphor, and its careful consistent
extension to exploit a computer graphics implementation.” However, the editors’ workbench will become
more than a metaphor as we aim to build a system that on the functional, hardware, and almost haptic (i.e.
hands-on, touchable) levels reproduces, augments, and organizes work with manuscript images into
image-based editions.
TEACHING AND LEARNING
As in any endeavor of developing a complex system the knowledge, creativity, and skills of the people involved are critical to its ultimate success. Students at various levels of their studies will be participating in
the project, and thus will bring with them different levels of skill. To accommodate and integrate students and
researchers from interdisciplinary environments, and to ensure that the architecture is flexible and open for
creative solutions and responds well to potential errors, unavoidable in any learning process, EPT will use the
end-to-end skeleton system advocated by Harlan Mills. In this approach the system grows from a basic loop
with stubs (subroutine calls) by adding functionality to gradually increasing components. The original
skeleton is developed by the most advanced researchers. This approach will help maintain the correctness of
the system throughout the entire process and at the same time will separate and divide design and
implementation tasks among different groups of students. The encapsulation and implementation of Object
Oriented Programming allows developers to define self-contained modules such as parsers, search facilities,
and image manipulation tools, while limiting interaction between separate components to interfaces, thus
reducing the probability of errors. Ultimately all these modules are plugged into the skeletal system and
become available through a portal to serve the user.
We will illustrate this approach with an implementation of the customizable graphical search tool
(CGST-XML) for XML documents that is a part of the development for ARCHway. CGST-XML is an
application built around the Model-View-Controller design pattern, wherein a user is provided with a number
of options (buttons) to control the operations of the system. Specific to CGST-XML is that the operation of
the system is organized into three steps: preprocessing, action, and post processing. Each step can be executed
by a number of modularized components that are available to the user. Through dynamically loaded
configuration files, the user not only can decide what particular programs, tools, and implementations are
used, but also can change the Graphical User Interface to one that best suits the current stage of the editorial
process. Additional implementations of separately developed modules can be added to expand the system,
multiplying the functionality of the system. The system is reminiscent of a plug-in oriented Web browser, but
instead of selecting specific applications based on a suffix, users have the ability to create their own suites of
operations. As a specific example, we will show CGST-XML in action with several possible configurations to
demonstrate the extensibility of the project.
REFERENCES
F. P. Brooks. The Mythical Man-Month: Essays on Software Engineering, 1995.
M. S. Brown, W. B. Seales, K. Kiernan, and J. Griffioen. “3D Acquisition and Restoration of Medieval
Manuscripts.” Communications of the ACM: Special Issue on Digital Libraries, May 2001.
E. Gamma, R. Helm, Ralph Johnson and John Vlissides, Design Patterns: Elements of Reusable
Object-Oriented Software, Addison-Wesley, 1996.
Handbook of Discrete and Computational Geometry, edited by J. E. Goodman and J. O’Rourke, CRC Press,
1997.
A. Hunt and D. Thomas, The Pragmatic Programmer, Addison-Wesley, 1999.