April 2004


T-Space introduces massive digital storage, access and convenience

Build it and they will come. This isn’t about some Hollywood fantasy. It’s about T-Space: a new state-of-the-art service at the University of Toronto for preserving and distributing digitally formatted research and education material such as high resolution images, complex data sets, learning modules, documents, audio and video.

T-Space is a practical, efficient, user-oriented service that will revolutionize access to and preservation of research and education materials in Ontario. The space has been built. But will they come?

Building the content of high-tech repositories such as T-Space can be a very slow process. It’s usually easier to sell technology if you can show potential users a demonstration project with some content as opposed to a black box, explains Rea Devakos, T-Space service coordinator.

For example, in technical terms, T-Space is certainly very impressive. It uses an IBM P670 server, a SAN main storage running two FAStT 500 storage servers and a backup Tivoly Storage Manager 4.2, tape library with a 100-terabyte (TB) capacity. Combined with the user-friendly, feature filled, almost idiot-proof program interface based on MIT’s D-Space open-source technology, UofT has built a very impressive service.

What the technical information doesn’t tell the faculty departments and instructors, who Devakos hopes will populate T-Space with their material, is that T-Space provides them with a level of convenience and visibility that they could never achieve on their own, with their own websites.

Where visibility is concerned, for example, T-Space has already attracted the attention of Google. “Our site is Googled weekly and T-Space content has been virtually assured a prominent ranking,” Devakos points out enthusiastically, noting that Google normally only crawls sites monthly.

Also, with the university’s gigabit connection to ORION through GTAnet, once T-Space becomes linked with other research and education institutions, the data will be far more readily searchable than individual faculty or department websites.

As of the end of March, T-Space was populated by a respectable 1,905 files submitted by eight UofT ‘communities’: G7/G8 Research Group, Health Sciences Information Resources (FIS 2135), Institute of Women Studies and Gender Studies (IWSGS), Knowledge Media Design Institute (KMDI), Munk Centre for International Studies (MCIS), OISE/UT, Office of International Surgery (OIS), and the University of Toronto at Scarborough.

“Right now we’re using only about two percent of existing space,” says Gabriela Mircea, Project Programmer with the university’s Information Technology Services. The storage capacity is now 9 TB and by 2005, the capacity will be 72 TB.

Both Mircea and Devakos are confident that the database will increase once the word gets out about T-Space and its many benefits, such as “tiered-access” which currently enables members to restrict access to the content they put on T-Space. Currently access can be restricted to users with UofT IPs, but ways of narrowing the level of access to specific IP addresses or through the use of passwords is being investigated.

T-Space, like Google, is a custom portal. Faculty members can embed a search within their web pages, which will allow each document to become searchable. Furthermore, each document will get a persistent URL or identifier that will not change. Anyone who uses the Internet for research knows the inconvenience and frustration that goes with URL’s that tend to change all the time.

Because T-Space provides services that make the design, loading and maintenance of content on their site so convenient and easy, faculties don’t have to spend their resources trying to design and develop their individual sites. And because T-Space is designed to last as long as UofT has a library system, faculty members have the peace of mind in knowing that their work will also be preserved at least as long as UofT has a library system.

Because Devakos and Mircea are determined to keep the files safe and accessible “500 years from now”, there are still some decisions that have to be made about how, in the long term, information is preserved. While html and rtf coding, for example, are believed to be the most durable format because they are open source codes, digital preservation techniques for proprietary software and complex objects are still developing.

“So far, we’ve received mainly text-based files but we have the storage capacity for much more,” says Devakos. “We’re getting some videos, and we’re working on preparing data sets to be uploaded, but clearly we’re expecting the volume and size of files to increase considerably.”

More information is available at the following three web sites: http://TSpace.library.utoronto.ca, http://www.dspace.org and http://libraries.mit.edu/dspace-mit/news/dspace-news.html.


Back to Headlines