Computing & Information Services

Murdock Open Source Initiative

Funded by the M. J. Murdock Charitable Trust and Reed College
April 2007 - December 2009

Background

Reed's technology environment has been enriched over the past decade with hundreds of databases, software packages, and other digital materials to support instruction, research, and administrative activities. The complexity of networked resources has serious implications for both usability and sustainability. As these resources become more numerous and diverse, they become increasingly difficult to use and expensive to maintain. The task of adding new software packages and enabling them to interact smoothly with one another becomes more formidable as well. 

This project focuses on ways to rebuild internally developed software with mainstream open source technology, develop new software with open source tools, and integrate internal, commercial, and open source software in a single framework that is easy to use and cost-effective to maintain. Our hope is that these efforts will produce a robust set of networked resources customized to the needs of the college and easy to modify in the future.

Approach and Staffing

This project involves three dedicated positions: a web programmer to work on user interface and web application development, a system programmer to work on open source software implementation, customization, and system integration, and an instructional technologist to provide documentation, training, and consultation for faculty and students in the use of the courseware toolkit

Web programmer –– The web programmer is responsible for re-building existing web applications and developing new academic and administrative applications using PHP, Pearl, Python, Ruby-on-Rails, and other open source tools. The web programmer is also involved in building the front-end of the campus portal and extensions for the courseware toolkit.

System programmer –– The system programmer is responsible for integrating commercial, open source, and internal software, back-end development of the campus portal, and implementation of the courseware toolkit, the identity management system, and the conduit between the digital collection and the courseware toolkit.

Instructional technologist –– The instructional technologist is responsible for assisting faculty and students in the use of Moodle and other instructional tools, preparing and disseminating documentation, conducting outreach and training programs for faculty, identifying and writing specifications for Moodle extensions, and serving as the liaison between faculty and programming staff during the development and testing of instructional tools.  

Project Objectives

The project involves several different types of specific technology objectives including:

  • re-implementation of internally developed systems;
  • deployment and customization of new systems;
  • integration of internal, commercial, and open source software.

Re-implementation of existing systems

In order to make internally developed software packages easier to use and more sustainable, we are redesigning the user interfaces and changing the underlying program code from Tango, a commercial web database language, to open source technology including Ruby-on-Rails, PHP, Perl, and Python. Among the specific systems we are rebuilding are: the registration system (SOLAR), the student information system (Gateway), the alumni information system (Reedlink), the campus-wide file system, and a variety of web and database applications in areas such as student orientation, career services, health center records, budget monitoring, student tutoring, and computer account management. 

Deployment of new software

New software implemented within this project includes two foundational systems and a number of smaller academic and administrative applications.  The two foundational systems are a courseware toolkit and an identity management system.

Courseware toolkit –– Acourseware toolkit is a suite of software programs that faculty and students can use in their teaching, learning, and research activities. Reed's toolkit is based on Moodle, an open source package launched in 2002.  Moodle has features that support the creation of web syllabi, electronic submission of assignments, distribution and analysis of online quizzes, surveys, and self-paced exercises, creation of class email lists, blogs, wikis, and online forums, e-reserves, grade keeping, and shared use of copyrighted digital resources. In addition to implementing and customizing these features, Reed is creating new academic software modules in response to feedback provided by faculty and students during the pilot phase. These modules will be added to the library of Moodle extensions so they can be used (or adapted for use) by other colleges in the Pacific Northwest and elsewhere.

Identity management system –– A campus identity management system is a set of tightly linked tools that perform three basic functions: (a) identification: a directory storing personal and role information (such as faculty-student-staff status) for all individuals who access institutional computing resources; (b) authentication: login methods, such as passwords or biometric tokens, that are used to confirm that a person is who he or she claims to be; and (c) authorization:  the process used to grant permission for specific users to gain access to specific resources (such as student records).  

A unified identity management system enables users to have access to a wide variety of software, databases, and digital materials by means of a single sign-on to the campus network. It permits individuals to update personal information in one place and have that information automatically updated in all college systems. And it helps to address security requirements by insuring that resources, such as confidential databases, are accessible only to authorized individuals.

Reed is building an identity management system, customized to the specific needs of the college, based on internally developed software and industry standard open source utilities such as LDAP and Kerberos.

Integration of internal, commercial, and open source software

The most important aspect of the project is to integrate all of the software components into a unified framework. Integration is taking place at the user interface –– the front-end or portal of the computing environment –– and within the underlying technology, the back-end. Front-end integration includes a single sign-on feature that allows users to access the courseware toolkit, the student information system, the digital collection, and many other resources. It includes standardized navigation controls and has web page views that are customized to the user's group (faculty, staff, student, alumni, etc.) and to the individual user (for example, students can access specialized resources only for courses in which they are enrolled). The goal of the front-end integration is to make networked software, databases, and digital materials as easy as possible to access and use. 

Back-end integration includes linking Reed's web applications to the identity management system, creating conduits between the courseware toolkit, the campus-wide file system and other key resources, and connecting databases and web forms to one another via open source tools. The goal of this integration is to make resources easier and less expensive to manage, upgrade, and expand. 

One of the most important integration tasks is creating a conduit between the courseware toolkit and the digital collection we developed with support from the Keck and Booth-Ferris foundations. The software we selected for the digital collection is CONTENTdm, a commercial system. This software enables members of the Reed community to access and search a rich collection of images and other digital objects from anywhere on or off campus. By establishing a conduit between the digital collectionand Moodle, we have made it easy for faculty to use a wide range of digital objects when they prepare their course materials.  We also created a link between our web authentication software, CoSign, and Moodle so that Reed users and colleagues from other institutions can access the courseware easily from on campus or off.

Timeline

Spring 2007

Begin project staff recruitment
Purchase and deploy administrative PHP servers
Purchase and deploy courseware toolkit and ID mgt. servers

Summer 2007

Appoint project staff
Begin re-implementation of student & alumni information systems 
Begin implementation of courseware toolkit and ID mgt. system

Fall 2007

Connect identity management system and courseware toolkit
Release courseware toolkit for faculty and student use
Begin development of portal and administrative web applications

Winter 2007

Complete re-implementation of student information system
Complete re-implementation of alumni information systems
Begin implementation of campus-wide file system (AFS)

Spring 2008

Release campus portal (IRIS)
Submit Year 1 interim report

Summer 2008

Begin integrating digital collection and courseware toolkit
Continue development of administrative web applications
Implement courseware toolkit extensions
Complete implementation of campus-wide file system (AFS)

Fall 2008

Release general documentation for digital collection and courseware toolkit
Continue development of courseware toolkit customization and extensions
Begin re-implementation of minor administrative applications

Winter 2008

Begin re-implementation of student registration system (SOLAR)

Spring 2009

Continue implementation of identity management system
Submit Year 2 interim report
Begin 5 year strategic planning exercise and project retrospective

Summer 2009

Continue re-implementation of minor administrative applications

Fall 2009

Complete re-implementation of student registration system

Spring 2010

Launch 2010-2014 five year technology cycle

Summer 2010

Project concludes -  submit final report

Outcomes

When the project is completed we expect to have:

  • a robust set of electronic tools, based on the Moodle learning management system with Reed-specific enhancements, that faculty and students can use to facilitate their instructional, research, and other collaborative activities;
  • a variety of new web-based utilities to support administrative staff in admission, financial aid, student records, alumni affairs, career services, health and counseling, and the business office;
  • an internally developed, open source portal that will enable all members of the Reed community to have single sign-on access, simplified navigation, and easier use of a wide range of databases, software packages, and the digital collection;
  • a software environment, based open source tools, that is significantly more sustainable, cost-effective, and expandable than our current environment;
  • a set of networked resources that are highly integrated, secure, and yet sufficiently flexible to accommodate growth.

Project Significance

In the near term, the project is adding a variety of new resources to the network while reducing the complexity of signing on, locating, and using those resources. Enhanced function, improved ease-of-use, and tighter integration of our software environment will allow all members of the Reed community to take greater advantage of the benefits of new technologies.

The long-term significance of this project is that it will enable the college to continue to provide a robust and growing set of technological resources to its community while constraining future costs.

Three items in technology budgets whose rapid growth threatens the ability of small colleges to sustain and expand their electronic resources are labor costs, software license costs, and network security costs. 

  • The technical complexity resulting from the deployment of many new software systems each year creates pressure to increase the size of the technology support staff.  Unless a college is willing to accept a less stable and less usable technology environment, this pressure inevitably leads to more staff and higher operating costs.
  • Commercial software license cost increases frequently outpace inflation, rising by double and sometimes triple-digit amounts in a single year. 
  • Network security improvements, including identity management systems, necessitated by the rising number of viruses, worms, and other types of intrusions, have consumed 10% or more of an institution's annual technology budget in recent years and these costs continue to rise.

By basing Reed's underlying technology on open source products, and building rather than buying key components such as identity management systems, we will avoid the purchase and annual license fees of a number of commercial packages. By building new systems, and rebuilding older systems, with standardized open source tools, we will limit the amount of labor required for software maintenance and development, and thereby reduce the need for additional staff.  Undertaking this project will make Reed's investments in technology substantially more cost-effective for many years to come.

 

Progress Report – July 2008

Progress Report – January 2009

Progress Report – June 2009