Archon - model for unifying archival description and access

I've been working with Archon since last August, when I attended a demo and program on the topic at SAA 2006 in Washington. Inspired by the presentation, I installed and configured Archon 1.00 on my laptop (Archon is a PHP/MySQL/Apache-dependent application that runs on MS Windows, an approach similar to that taken with another recently-released archival application, Archivists' Toolkit, which requires a Windows OS and MySQL.) I recently upgraded to Archon 1.11, which I hope to test in the near future.

I've found Archon to be of greatest use as a working example of unified archival description and access. The application itself, at 1.00, was unable to support the load of large EAD documents (example, a 500 MB EAD XML file with a lengthy component list did not load successfully into Archon). Additionally, the entire concept of exploding XML content into the table/columnar structure of an RDBMS is flawed (see earlier post). XML is a great structure for defining the box/folder/document structure of an archival collection; why undermine it by moving its content into MySQL?

With this background in mind, I read a pre-print of an article on Archon (which I found online, but will not link to, at the request of the authors, who are members of the Archon team at UIUC). The article defines the goals of the pilot project that led to the creation of the Archon software: creating a web-based application that enabled the creation of EAD by archivists and staff lacking a detailed knowledge of the standard. The RDBMS implementation decision is discussed in brief; several reasons are listed or implied, including the easy ability to integrate authority control features and the ability to reduce development costs.

Archon contains an automated installer and the authors note the importance of this installer and the open source, freeware nature of the supporting tools (Apache, MySQL, PHP). Configuring PHP on Windows and installing Archon, even with the installer, are not trivial technical tasks. Thus, the idea that the tool can be employed by financially-strapped institutions lacking technical support may be a bit of a stretch. For example, with the upgrade to 1.11, I found that I had to locate and launch a specific PHP script manually (based upon a deduction, not specific instructions) for the process to successfully complete.

There are sections in the pre-print on the Archon API. If you use the Archon tool, make sure to carefully review the license agreement, which I heard described at the SAA conference. My interpretation of Archon's 1.11 terms: No derivative works outside of your local archival or personal use without UIUC permission; modification of Archon's code for commerical purposes requires permission.

In summary: Archon's greatest value, beyond the goals defined by the authors in the pre-print, is to serve as a functioning model for unifying archival description and object access in a single application. I need to test 1.11 to check the component list size problem/XML file size problem that I described above.