DARE Project Chronology
Starting the Challenge
Cream of Science is 'invented' during a meeting of the DARE partners in the summer of 2004, while discussing different strategies to increase the volume of DARE repositories and to increase the awareness of scientists.
The Cream Project starts with preparatory work done by DARE programme management. This results in a planning document that is sent to the DARE partners who quickly respond with their commentary. It becomes clear that the actual work will be delegated to the university libraries. However, they do not have the manpower to deal with it and fear that there is not enough time to get the work done. It also becomes clear that most people expect copyright to be the major problem for Cream. The initial enthusiasm for Cream is now overtaken by scepticism.
The DARE librarians make several decisions. They will aim for a complete publication list (metadata) for all Cream scientists, but not for full digitisation (digital object files). The project deadline is moved from January to April. They will implement two sets in their repositories: one for Cream and one for DAREnet (the DAREnet set already exists and is 100% digital and open access).
DARE programme management decides to create support teams to assist the libraries in the work that needs to be done: a metadata input team, a scanning team, and one person to find publications via inter-library loan.
At this time the three major challenges are:
- myths about copyright
- lack of manpower
- no clarity about the workload for Cream (total number of publications, work to be done for scanning, metadata input and inter-library loan)
In fact, it becomes clear that university libraries only have limited figures about the volume, publication, storage and other facts regarding the scientific output of their own scientists. And most of this output is not part of their catalogue and local digital storage.
While the libraries work hard to select ten prominent local scientists for Cream, and to get some hard data on their publications, DARE programme management starts work to support the libraries. This results in three separate teams:
- a team in Groningen University for metadata input (mostly local students)
- a team for scanning (contracted to Strata Preservation bv)
- a person to handle ILL (inter-library loan) in Wageningen University
Later (in December) another team for metadata input is added (contracted to DMP bv).
Due to the fact that hard data is not available, estimates are used to determine the size of the teams and the volume of the contracts. The estimates are as follows:
- 150 scientists in Cream
- 300 publications per scientist = 45,000 publications in total
- 50% already available in digital form = 22,500 publications to be scanned
- 10 pages per publication = 225,000 pages to be scanned
- 25,000 metadata records to be keyed in
At this point in time, it is expected that most libraries will use the support offered by DARE and only two libraries clearly indicate that they will do all the work locally. Logistical issues regarding scanning are discussed. Transport, turnaround time, demarcation of start page and end page, quality, advertisements etc. The most important issue turns out to be the local policies that do not allow journal volumes to leave the building. The 'classic' library rules prove to be very strict. Because of this consideration, another four libraries later decide to do the scanning locally and not use the scanning capacity offered by DARE.
New questions arise. Do we include monographs and articles published in popular (non-scientific) magazines? How to solve the issue of variations in author names? How to implement the two sets? Is not selection of the Cream publications the task of the service provider, in this case DAREnet, so why do we need sets anyway?
Libraries start to send their first publication lists.
A Cream of Science Newsletter is created to enhance communication around the project.
The Cream idea turns out to be more successful then expected: 173 scientists want to be included and the hoped for 'me-too' effect becomes visible. SURF decides to create additional funding for Cream and PICA decides to support Cream with a special feature in their inter-library loan system (at no extra charge). Metadata issues arise around the correct use of Dublin Core for journal articles.
January - April 2005.
Cream has now become top priority in most university libraries. More and more hard data about volumes and workload becomes available. More scientists want to participate: 207 in total. In some cases, an entire faculty wants to participate. A waiting list is created.
By the end of January, most of the initial problems and uncertainties seem to be resolved, although libraries are a little frustrated by the fact that DARE programme management is forced to change the support planning frequently. The DARE teams are up and running, more and more librarians decide to devote additional manpower to Cream. Work for Cream becomes almost a part of daily life. In most DARE meetings, a lot of time is spent discussing Cream.
DARE programme management now has three major tasks:
- Daily co-ordination of DARE teams and local library work for Cream (trouble- shooting and monitoring)
- Development and testing of the central DAREnet applications (harvester and Web site)
- Preparation for the JISC/CNI/SURF conference during which Cream will be presented.
By mid-April, testing of the harvester begins. New metadata issues are discovered, OAI-PMH protocol implementation errors are corrected and problems with sets are addressed.
The DAREnet Web application is tested and discussed with the DARE libraries. Although a richer functionality was planned, important features like full-text indexing and drill-down have to be postponed until after 10 May, the launch date. Too much time is needed for testing the harvester and trouble shooting the DAREnet Web application. Even though there is no formal planning and some people are getting frustrated by the continuous changes, the DARE community works very, very hard, with a lot of enthusiasm and dedication. Cream of Science is a catalyser.
Cream is doing overtime. The launch date is just days away. Harvesting errors are almost a daily routine. A lot of work is still being done on the website. People are getting excited about the presentation and the surprise celebration. During the final days, everything seems to be working well and looking good. When the big day arrives, everybody is nervous, but confident that they have done a good job; of course there is still room for improvement, but they have done their best. The outcome is satisfying.