Last week I presented on “Cyberinfrastructure: Reflections from TAPoR to Tools at the Canadian Research Knowledge Network Annual General Meeting 2008 (they have a PDF of the slides.) I was part of a panel on cyberinfrastructure that included an interesting presentation by Walter Stewart of CANARIE who made the point that the big issue is people. While many still don’t have access to the technical infrastructure that would facilitate their research, the big challenge is professional staff/collegial support for digital research. If one looks at the life span of a typical project one can see where people are needed:
- Conception: when colleagues in the humanities are imagining a project that might have a digital component they need good advice.
- Application: if they go forward with a grant application they need help articulating the digital component so that it is clear and technically accurate.
- Modeling: if they get a grant they need help training the students who do the work, they need help making the technical decisions that affect downstream research, and they need help managing the implementation. Most colleagues don’t have the experience needed to bring a digital project to completion within budget and on time.
- Virtualization: most digital humanities projects go out to the web and projects need help delivering them to the web and virtualizing the service so that it can be maintained as a stable machine. Typically a project will get funding to pay for the programming needed, but not for ongoing maintenance. We have found that one way to stabilize a project so it doesn’t need constant updating is to create a virtual server with all the layers of applications (lets say a certain version of Ruby and MySQL) frozen so that updating something on the server doesn’t break the service. This takes professional server support that is ongoing so these projects can be migrated from machine to machine over time.
- Maintenance: even virtualized projects need occaisional maintenance if bugs are found or if new data needs to be added. If the programming was done by a graduate student who has long since gone, as is usually the case with grant funded projects, then the cost of maintenance can be exorbitant. The solution is not to use only professional programmers as work on projects is one of the best forms of apprenticeship in the digital humanities for graduate students. What we need is permanent programming staff who oversee digital projects, guiding the graduate students, and making sure that code is documented so it can be maintained. These project manager level programmers then provide the long term knowledge so that a new student could be hired to fix something and guided around the project.
In short I think we can begin to articulate a baseline of cyberinfrastructure and support needed at research-intensive universities to support a culture of digital humanities projects:
- Servers: Research-intensive universities (RIU) need to run flexible servers capable of hosting the development and deliver of projects. These need to flexible in the sense that service models that limit service to specific applications (we only support PHP) almost always fail to evolve at the speed of projects leading projects to spin-off their own servers outside the support umbrella.
- Labs: RIUs tend to see a proliferation of labs “owned” by particular projects. Given how most of us and our students have laptops we no longer need labs specifically for work. Instead labs are becoming places for access to specialized tools (large format scanners, special software, and visualization displays) and places for social research. Labs, in my experience, are becoming places where people work together whether meeting over an interface or testing a project. I would argue that labs should be “socialized” and brought together so that projects share space so they can learn from each other. That said, labs are still needed.
- Project Managers and Technical Staff: Most important, following the outline of how knowledgable people are needed I would argue that RIUs need to have a mix of technical staff with project management experience to guide projects through from conception to long-term stability. Such staff can be in the library, faculty or computing center, but they should be coordinated. These staff do not replace the grant funded people brought on to work on projects, but they provide the advice to get the grant and oversight to manage contract staff.