Elizabeth Franklin, Celeste Vallance and Libor Coufal are members of the Digital Preservation team at the National Library of Australia.
On World Digital Preservation Day 2025, we published a blog post for the Digital Preservation Coalition titled We Will WARC You: preserving access to published CD-ROMs at NLA. In it, we described our novel use of WARC technology to provide access to complex, HTML-based publications from our collections — not harvested websites, but offline publications containing thousands of interconnected files in a wide range of formats.
This led us to an intriguing question: Could WARC technology also be used to provide access to other complex digital objects whose functionality depends on the relationships between their files?
Onward from Phase 1
In Phase 1 of the project, we focussed on generally less-complex HTML publications. We identified 515 items as in-scope for this phase. All were transformed and ingested between August 2025-April 2026. As our experience grew, we started thinking about what could be included in Phase 2. As well as more complex HTML publications requiring additional attention, we began considering other complex digital objects whose functionality had proven difficult to preserve. We already knew that we had PDF files in our preservation system that contained links between them and were meant to be browsed as a whole publication. Although they rendered as standalone files, the user experience was incomplete without their interactive functionality. At the time of processing these carriers into our preservation system, we had identified the linkage issues but had no solutions to the problem, so accepted the reduced functionality as an unavoidable compromise.
But now, with new tools and experience, we asked ourselves: if the linked PDFs conceptually behaved like HTML publications, could WARCit also handle them?
Beyond HTML
The answer, it turned out, is yes.
By adapting our existing workflows and developing new Python scripts, we were able to restore navigation between linked PDFs, allowing users to browse them as interconnected publications rather than isolated documents.
Figure 1 – Conference proceedings designed as a set of inter-linked PDF documents accessed through an interactive navigation page
As we delved deeper into the collection, we encountered more challenges that looked solvable with WARC: interactive PowerPoint files, widget buttons, embedded multimedia and executable title screens. We found ways to adapt their functionality and navigation to work within the restrictions of the ReplayWeb.page renderer. While the technical details differed, many of the same principles that underpinned our HTML workflow could be applied to these publications.
Figure 2 - Interactive slideshow with optional audio track
Figure 3 – PowerPoint-based publication linking to files in other formats
Figure 4 – PDF publication with navigation and embedded media
Figure 5 – Searchable index of publication linking to hundreds of issues and articles
These challenges also required us to think beyond technical solutions. We aimed to replicate all features present in the original publications, but this turned out to not always be possible within the WARC format. This meant we had to consider what was essential to the experience of the publication, and what could acceptably be compromised on for an access copy.
Progress and lessons
Since we began this work in August 2025, we’ve transformed 763 carriers in total, representing over 626,000 files in 185 unique file formats.
We did all this while balancing our BAU work and other demands.
The most significant achievement of the project has been turning the knowledge and experience gained from individual publications into repeatable workflows. As patterns emerge across hundreds of publications, processes that once required significant manual intervention were largely automated. This does not reduce the need for specialist expertise, but allows that expertise to be directed towards genuinely new challenges rather than repetitive processing tasks.
The result is a cycle of continuous improvement. Lessons learned from one publication help us tackle the next, and publications initially set aside because they appeared too complex often become manageable once the necessary tools and experience have been developed.
Looking Forward
What began as a way to provide access to Offline HTML publications has evolved into an approach that can be extended to a much broader range of born-digital publishing formats. Although we have more publications set aside for a possible Phase 3, we are taking this opportunity to reflect on what we have learned and explore what else might be possible. And if the past year has taught us anything, it's that the next opportunity might be hiding in a format we haven't thought to WARC yet.