Technical aspects of the Online Edition

1. Static site created by HCMC in 2026

This online edition of the Cambridge Edition of the Works of Ben Jonson (CEWBJ) is an entirely static re-implementation of the original site. The re-implementation was done by the University of Victoria Humanities Computing and Media Centre (HCMC). The original web application, created by a team at the Department of Digital Humanities (DDH), King’s College London, was first scraped, and then a number of remediation steps were taken to create a standalone static site requiring no back-end server infrastructure:

• The HTML pages were heavily remediated to make them valid and well-formed XHTML.
• Listings pages which originally rendered items 20 at a time were combined to create one-page complete listings, for speed and convenience.
• Each of Jonsonʼs works was converted from a multi-page rendering into a single page containing the entire work.
• Server-driven functionality was replaced by client-side JavaScript where necessary.
• Subcollection-level filtering functionality was amalgamated into a single search page allowing users to search the entire collection.
• The Chronology visualization was converted to a simple page listing the events.
• Redundant JavaScript libraries were purged.

The objective of the HCMC team was to remove complexity and dependencies, while preserving as much as possible of the original material from the site, along with its basic design features, and to add a more robust search facility. The resulting site should be resilient and maintenance-free. Meanwhile, the Works are being migrated into the Linked Early Modern Drama Online (LEMDO) platform, where they will be republished in due course.

The technical description of the original Kiln-based publication framework is described below, as it appeared on the KCL-hosted site.

2. Original publication engine created at King’s College London

The major ground breaking accomplishment of the Cambridge Edition of the Works of Ben Jonson (CEWBJ) has been the development of reusable digital infrastructure for future scholarly publication, involving the handling of highly disparate source materials, and producing an end product of saleable quality whilst ensuring academic credibility.

The use of this tool for other such projects would enable a far more rapid delivery, and a lower level of technical knowledge. It will provide a foundation to go forward with such developments, delivering enhanced facilities to handle digital editions, multiple version of content, facsimile contents aligned to text, and different styles of editorial intervention.

Predicated on the prevailing standards relating to the archival representation of text (TEI XML), this will facilitate and indeed encourage working to a best practice standard, for CEWBJ and future similar projects.

In the course of working on CEWBJ the xMod publishing library – previously developed at DDH and the technical basis of the digital edition – has received extensive changes and enhancement and is now a considerably more sophisticated and powerful tool with the potential for very broad application across the broadly conceived Digital Humanities. Such was the significance of these changes, rather than continue to increment a version number, xMod was renamed Kiln in recognition of the substantive nature of the developments that have been made. Key developments have been as follows:

• Kiln is now Open Source (published on Github at http://kcl-ddh.github.io/kiln/) and fully documented on ReadTheDocs (http://kiln.readthedocs.org/en/latest/).
• CMS integration and 'headless' mode: Kiln is able to act as a ‘content provider’, integrating familiar authoring tools for creating standard web content, in addition to its existing capabilities as a content publishing framework for complex XML documents. In this ‘headless’ mode Kiln serialises content and passes it to a chosen Content Management System for rendering.

For Jonson, this works in the context of the Django CMS Mezzanine, but this approach can be adapted to work with most current CMS systems. A current project has successfully integrated Kiln with WordPress (for example).

• Indexing and searching capabilities are natively provided by Solr, which is integrated into the Kiln distribution.
• The Sesame triple store is also now integrated into the Kiln distribution by default.
• Kiln now integrates a powerful frontend templating mechanism inspired by the Django framework, which allows for better separation of concerns by splitting out frontend HTML templating from the underlying XSL logic. This allows for web frontend designers to work on Kiln templates without XSL knowledge.

In addition to Kiln, the Jonson project has significantly contributed to two further Open Source tools:

• Bacalhau (published on Github at https://github.com/kcl-ddh/bacalhau) is based on the Castanet algorithm,¹ and automatically creates subject hierarchies using WordNet.

It has been used in Jonson for experimental visualisations during the development process and for the facetted browse functionality.

• Textviewer (https://github.com/kcl-ddh/textviewer) is a JavaScript-based dual pane text viewer with sophisticated functionality for synchronised scrolling which respects document structure, together with a number of end-user-customisable display options.

Following the publication of Kiln, Bacalhau, and Textviewer as documented Open Source tools, all are now being used independently of KCL’s Digital Humanities department.

The Online Edition web application is premised upon a workflow which allows the site to be continuously updated as required:

• This process begins with the use of automated NLP techniques to extract person, place, works, and date entity data from the project’s XML corpus.
• The corpus is then indexed in Solr, together with the extracted entity data, and categorised by document type.
• Once this has been done, the Solr index is used to generate any document or content listing screens (these are effectively all preset searches), and to create the browsing and searching interfaces.
• The XML content itself is rendered by Kiln using both the default XSLT for TEI P5 documents, and further XSL templates customised for the Jonson edition.