Digital Repository 2.0: Lessons Learned and Applied

Abstract

The Portal to Texas History at The University of North Texas Libraries is a comprehensive system for storing and providing access to a very large number of digital objects of historical significance to the state of Texas. The Portal is currently in its second iteration of development, and in this presentation we hope to examine some of the lessons learned from our initial efforts and how they shaped the decisions made in the current system and the future.

We will briefly overview the first system that was put in place, and goals that we had in mind for it. We will cover some of the ways that it was successful, and some of the limitations that we encountered.

The things that we will highlight about the former system include:

Data Model

Overview of the old format and limitations encountered. Technologies Utilized

Issues with products that have a small user and developer base. Our XSLT experience. Architecture

A look at the single-machine model and how it relates to scalability and redundancy. Development and Workflow

The “learning project management as we go” adventures. Next, we will cover the goals behind the current version of the Portal. As there are several aspects of the system, we’ll be looking at different areas of importance that drove our decisions.

The topic areas will parallel those covered in the previous system.

Data Model

The improvements made to our current data format, based on the limitations of the old format. Technologies Utilized

Technologies we used to build our new system. Reasons for selecting these specific technologies. A brief overview of the potential of web frameworks, focusing on Django. Architecture

Explanation of and benefits of the “shared-nothing” approach. Scalable and Deployable. Development and Workflow

Moving from one developer to several. Content management and workflow tools and procedures. Division of tasks and collaboration. The motivation behind “rolling our own”. Our goal in this presentation is not so much to present a “this is how everybody should be doing it” argument, but rather to highlight some of the issues that we encountered and our approaches to resolving them. It is our hope that other groups can learn from our mistakes and successes as they seek to implement their own Digital Repository systems.