>>> "B.K. DeLong" <[log in to unmask]> 11/19 9:27 AM >>>
At 08:18 AM 11/19/98 -0600, you wrote:
How have others dealt with the task of sorting through the tag set and
picking out what you'll use?
Being a relative novice at EAD I've dealt with this issue recently. You can start with the absolute minimum (to parse) required tags:
<archdesc level="collection"> [a level attribute is required]
You can also look at the EAD-lite tag set (it was part of the early EAD training materials).
However, the true difficulty in applying EAD comes in determining what you *specifically* want to accomplish with your SGML documents, and how you are actually going to do it. Are you tagging legacy documents, or creating new ones directly in EAD? Do you want the EAD documents to be in a searchable database, or just provide print copies? Is the database to be accessible via the web, or in-house? Perhaps you are starting with non-searchable, HTML versions converted from EAD instances. If creating a searchable group of documents, what kind of search engine is used, and how will it be configured? The granularity of your tagging is directly related to indexing. You may want to provide immediate but basic access to your information, and forego the detailed tagging process.
Anyway, it's a complex process. We've started with the ultimate, end-goal (providing a searchable database over the web), and are working backwards. We're tagging legacy finding aids for the most part, which involves a different type of decision-making. We've had to 'fit' the existing document into the tags, sometimes reformatting it (old finding aids sometimes have all kinds of information lumped together). Because finding aids are so inconsistent, and each institution's practices vary, it's been difficult to get simple answers to such questions. Each situation is different.
I too look forward to the Application Guidelines when they become available. For now, I've spent a lot of time working this out with colleages both in my own repository, and through this list, SAA meetings, and web sites such as the EAD Roundtable. All have been of great value. Good luck!
Katherine A. Hayes, Assistant Archivist
Niels Bohr Library, American Institute of Physics
(301) 209-3179 [log in to unmask]