You are on page 1of 8

Deliverable 2.

2
Publishing and disclosure systems
1. V2_Archive online portal
In the summer of 2003, V2_Archive has launched its public online archive portal1. The new
archive portal was built in continuation of the website for the DEAF03 festival, which already
offered a 'sneak peek' at V2_'s archive content.
The archive portal offers a public view on the documentation and context of V2_'s activities
from 1993 till present, including people, organizations, artworks and events that have played
a role in V2_'s history; documentation includes a collection of about 15,000 photographs and
a growing number of digitalized video fragments, in addition to a large number of texts,
images and links. The information visible in the archive portal is only a portion of the data
available in V2_'s archive database; large parts of this data collection are still available for
private use only2.
The interface of the archive portal was designed for a general audience; specific interfaces
for alternative visualization and mapping, and for selected target groups, will be researched
in the future.
A series of screenshots of various interfaces of the archive portal can be found in Appendix
1.
Functionalities of the archive portal include:
- Node pages and leaf pages
For every person, organization, event, artwork or keyword with sufficient metadata
and documentation in the database, a separate so-called "node page" is created on
the fly. The node page primarily groups and sorts links to other related objects in
V2_Archive. Pieces of documentation (articles, images, video) about the item which is
focused there, are shown in a pop-up window (a so-called "leaf page") when clicked.
- Indexes
Indexes ordered listings are available for people, organizations, events, artworks,
articles and keywords. These are the top categories in the archive portal. Each index
can be narrowed down to subcategories of these top categories. They can be sorted
alphabetically, by country and by date where applicable.
Furthermore, for most indexes (person, organization, event, artwork) it is possible to
display each object's metadata in minimal, default and extended view, and to toggle
each object's thumbnail picture on and off.
- Filters
Domains are information subsets of the object cloud in V2_'s archival database, in
most cases related to years, larger activities (e.g. DEAF festivals) or departments of
V2_Organisation (e.g. the publishing department). While browsing or searching the
archive, a visitor can choose to filter information to a specific domain.
- Intelligent search mechanism
The archive portal includes an intelligent search mechanism, created with the help of
the open source search engine ASPSeek3. The search results include not only
1

V2_Archive portal. 2003. V2_Organisation. 31 December 2003 <http://archive.v2.nl/>


More information about the content of V2_Archive and about earlier developments can be found in
Fauconnier, Sandra, Anne Nigten and Boudewijn Ridder: V2_Archive: Archive of living actualities.
2001. V2_Organisation. 31 December 2003 <http://lab.v2.nl/home/_docs/v2_archive.pdf>.
3
http://www.aspseek.org/
2

V2_Organisation Capturing Unstable Media

immediate hits related to the search query, but also suggestions of related objects
from the archival database
Keyword browser (tree)
A first specific information visualization application is the interactive tree of keywords
that can be browsed on the archive portal. This tree view shows the hierarchy of
keywords in V2_'s thesaurus and allows users to discover the thesaurus' structure in
a simple, interactive manner.
Relation models
The texture of the archive portal is defined by XML documents specifying the clusters
of relations that should logically constitute a "node". Such "relation models" have
been defined for actors, events, artworks, keywords, articles and media.

V2_'s archive portal was designed in such a way that all the dynamically generated pages
are indexable by Google and other major search engines. This was achieved by excluding
special characters (?, &, =) from the individual node and index pages' URLs. In this way, it
was avoided that the contents of the database would become part of the invisible web4.
This effort, together with a targeted e-mailed announced, resulted in rather high statistics on
unique visitors and page views; by December 2003, the archive portal received an average
of 47 unique visitors per day.
A variety of technologies is used to construct the archive portal, including:
- MMBase5 (an open source CMS);
- Arborealis (an open source project to transform XML to Java objects);
- a java servlet to cluster relations into 'nodes' using relation models;
- a graphics servlet based on ImageMagick6 for transforming images on the fly;
- an ASPSeek servlet for search engine indexing;
A generic "XMLServlet" written in Java acts as an interface between these different
technologies. Python communicates with this XMLServlet for various queries and transforms
the resulting XML to HTML using XSLT. For performance sake, most of these requests have
been preprocessed and cached, resulting in a static XML backend.

See the glossary for a definition of the Invisible Web.


http://www.mmbase.org/
6
http://www.imagemagick.org/
5

V2_Organisation Capturing Unstable Media

2. V2_Archive based on XML


Until the beginning of 2003, V2_'s archive database has been dependent on various open
source data management solutions, most notably the MMBase content management system
and the MySQL database application. In 2003, considerable efforts have been made in order
to migrate V2_'s archive content to XML (the so-called xmlbase), for various reasons.
-

XML-based archival content is available in plain text, a format unencumbered by


licenses or restrictions and suitable for long-term preservation.
The content is simultaneously human- and machine-readable.
The XML data are stored in a conventional file system with folders and files that can
easily be opened and modified.
The content becomes platform- and application-independent.
XML-based archival content can easily be transformed into other formats and makes
interoperability with other archives' content more easily possible.

For this purpose, an appropriate file structure and a series of XML Schemas have been
designed. Relations between database objects are described by URN's; these are resolved
to URL's or local paths by a small set of namespaces created for the archive.
For a snapshot of the file structure of V2_'s xmlbase7, please refer to Appendix 2. Near the
end of 2003, new editing interfaces were designed for xmlbase, based on technologies like
XSLT, Zope and XUL8. Here as well, a series of tools were designed in order to generate the
XML and to query this generated XML as if it was a database. The editor operates on RDF,
which is generated from the original XML backend using XSLT.
A screenshot of a preliminary design for the new archive editors, which work as an
application within Mozilla, can be found in Appendix 3.

http://v2archive.v2.nl/xmlbase see also http://v2archive.v2.nl/xmlbase/docs/200401_struct.txt


The development of the V2_Archive is open source. Mail archive@v2.nl to arrange personal access
to all projects through CVS or get more information about the technical issues involved.
8

V2_Organisation Capturing Unstable Media

Appendix 1
Selected screenshots from V2_'s archive portal, December 20039

Fig. 1. V2_Archive portal homepage

Screenshots were made in December 2004, in the Mozilla Firebird browser version 0.7, Mac OS X
10.3.2 (Panther).

V2_Organisation Capturing Unstable Media

Fig. 2. V2_Archive Node page of Event "Wiretap 3.05"

Fig. 3. V2_Archive portal chronological index of events: 1995-1996

V2_Organisation Capturing Unstable Media

Fig. 4. V2_Archive portal search results for query "robotics"

Fig. 5. V2_Archive portal tree view of keyword "engineering"

V2_Organisation Capturing Unstable Media

Appendix 2
Selection of file structure of V2_Archive's xmlbase

V2_Organisation Capturing Unstable Media

Appendix 3
V2_Archive xmlbase / XUL editors, preliminary design

V2_Organisation Capturing Unstable Media

You might also like