more at: version 1.1 demonstration

Just got out of a demonstration of the Archivists' Toolkit 1.1 beta. I have to say that this is an impressive tool from the processing side. Beyond applying AT, Lee Mandell, the Design Team Manager for AT, noted the need for a public access (web) component that's tightly coupled with the AT archival processing tool. After the session, I had a chance to speak with Lee briefly. Java development is required to extend the application. Having participated in the DSpace community for years, I'm aware of the difficulty of extending a Java-based application (as demonstrated by the number of DSpace instances running the canned interface and using the canned workflows).

AT and Archon offer a nice contrast, in terms of current archival tools. Judging by the interest in AT sessions at this meeting, Archivists' Toolkit has the most momentum in the archival community. I attended an Archon session at SAA last year. It has the more attractive architecture in terms of extension (being PHP-based). Archon's license terms (by my understanding, which prohibit redistribution after extension without UIUC's permission) perhaps turned off some developers. Clearly, the AT is the most full-featured and forward-looking tool in terms of archival processing. I know that after seeing the demo and listening to the speakers, I feel more informed about current archival processing practices and content standards. Also, there is clearly a strong base of support in the archival community for AT, which should translate nicely into community activity. Whether this is enough to overcome the architecture limit that I describe above...I don't know.

Then, for both AT and Archon, there's the issue of mapping XML to an RDBMS and how well or poorly that works. This is closely related to my current area of research. Given the issues described above, it's a downstream question, but one that needs to be assessed for each tool. The info that I have, based upon a quick conversation, is that there's a miscellaneous data field that holds EAD data that doesn't match to the defined tables/columns in the RDBMS; and, that there's no proven, OSS tool XML database tool that's analogous to MySQL in terms of user group support and reliability.