background image
The XML Family
of Technologies
58
form, but since the early 1980s audio
recordings of parliamentary sessions
have been kept, and an Access database
has until recently been used to hold
the essential metadata.The AMDA
project was initiated both to import
existing metadata into the new system,
and to create new metadata automati-
cally during the new digital recording
process.This leads to two distinct
benefits: the scope for error in the
ingest process is minimised, and the
metadata is stored in a much more
widely compatible format.
The fundamental technologies with which AMDA was put together are:
- Java 1.4, together with a Tomcat servlet container;
- A JSP/Servlet, with a Tomcat Servlet Engine for the GUI;
- JDBC 3.0, for database connectivity, together with JDBC Interface for data storage;
- Java API for XML Processing (JAXP) for XML operations;
- Oracle 8 DBMS.
The SIARD implementation partner Trivadis has provided some advice on technology
choices; this guidance had mostly to do with complex database issues.The SIARD team
defined the XML markup internally. In addition to this, the team examined other com-
parable solutions, including the US National Archives and Records Administration's
AERIC
81
and the Norwegian Arkadukt project.The primary focus of these projects,
however, had more to do with plausibility and integrity checks, which although
important are not the primary aim of SIARD.
The implementation of XML has been very successful in both the SIARD and
AMDA projects, and all anticipated benefits have been met. XML was selected for its
ease of editing and the fact that well-formed XML is automatically validated.The ways
in which XML can be transformed using (for example) XSLT also made it suitable for
these purposes. It should be stressed that neither project is completely finished yet and
there is much work still to do.
User and usability tests are under way for SIARD, and these should provide valuable
feedback for future versions. Another tool is in development that will facilitate the gener-
ation of metadata catalogues. It is planned that SIARD will eventually be extended to
include more specialised drivers for supporting database management systems other than
SQL Server and Oracle. AMDA is currently at the metadata ingest stage, and the user
interface will eventually be improved in accordance with preliminary user feedback. At
the time of writing (May 2003) the Digital Preservation Testbed
82
is testing SIARD
to verify its suitability as a migration pathway.
Schweizerisches Bundesar
chiv
AMDA: Debates overview
81
For more on AERIC, see Greg LaMotta's presentation from the ERPANET workshop in Bern, April 2003,
available online at http://www.erpanet.org/www/products/bern/bern.htm
82
http://www.digitaleduurzaamheid.nl
TWR2004_01_layout#62 14.04.2004 14:07 Uhr Seite 58