As I have mentioned it became clear early on that a database would be needed in conjunction with the webpage and images. I was, in fact, thinking about the database even before my independent study had begun. Karen Novick, of the SCILS Office of Professional Development Studies, immediately raised what would be the two main issues surrounding the database. Namely, what software would be used to create the database and what field names would be used.

The first thing to consider, in terms of the database, was what The Zimmerli was using and in this case it was Microsoft Access. As one would expect one record is created for each piece of artwork. In order not to duplicate efforts it was clear that it would be necessary to import data from The Zimmerli's database. Being that I have little background in the mechanics of database design I relied heavily upon The Knight in Shining Armor. It took the combined efforts of Wendy White, Gail Aaron (The Protector of Precious Images), and myself to provide him with the technical information needed to make an informed decision on how to proceed. Towards the end of the semester, Alex received the database records for all of the Mathers material as well as a copy of the database manual. Again, please refer to the Technical Paper: The Petra Project [/shadow/process/tech4.jsp] for more specifics.

In the meantime, the discussion of the appropriate fields continued intensely if not sporadically during the first year of the project. The final fields only being decided towards the end of the project. For me, the hardest part of this discussion was balancing my wildest dreams and reality. I would have loved to have a database and a search engine complete with Boolean search capability. Yet, the technical aspects of creating a search engine ultimately proved to be beyond the scope of the project. Could the database still be designed to allow for this in the future? I thought so. Again, a compromise was reached that allows keywords to be assigned for each image and perhaps one day the ability to search by keyword will be added. The fields that were ultimately used are a combination of the fields that The Zimmerli uses and those that have been used by other digital archives. The database created by The Visual Arts Data Service (VADS) in the United Kingdom [http://vads.ahds.ac.uk/index.html] was the primary model used. Like VADS there are two records for each piece of artwork: one for the original work and one for the digital archival master. In our case, most of the field names for the core record (ie. original artwork) are based on the original Zimmerli database and the digital archival master fields were agreed upon by the team. The field names for the core record are: Accession #, Creator, Medium, Date, Dimensions, Location, Rights, Title 2, Publisher, and Credit Line. The field names for the digital archival master are: Record ID, Description, Corporate Creator, Type, Create Date, Hardware, File Format, File Size, Resolution, Bit Depth, Color Space, Compression, and Rights. Finally, there is a field for keywords. The last part of the database to be completed was the input of the data after The Zimmerli's data had been incorporated into the database. Luckily, Alex was able to structure this so that I could input the data remotely and work from the comfort of my own home.