Daejeon, July 27–31
How do we prevent sysadmin and IT administrators from acting as the Grim Reaper of old digital critical editions? And who should be in charge of taking such decisions? We faced the urgency of this question in October 2024, when the 20-year span editorial work conducted by Eglal and Fekri Henein on the famous 17th century French novel L'Astrée was unplugged from Tufts University servers and left inaccessible for a few weeks. 180 html open-access pages, resulting from an invaluable critical work about what could be considered one of the most commented novels of the Early Modern Period in Western Europe, were on the verge of disappearing because of security breaches in PHP protocols. This is when Fekri Henein and Tufts IT services accepted to migrate the website to the University of Rouen’s servers. We then envisioned, together with the team of the Chaire d’Excellence en Édition Numérique, to convert the critical edition into XML-TEI, and to ensure the preservation of Henein’s work with the help of the Presses Universitaires de Rouen et du Havre.
The present paper proposes to present the various challenges we faced in trying to save this old digital critical edition and making it more sustainable, from dealing with Tufts IT administrators to retro-converting the edition and negotiating its integration within the Presses Universitaires de Rouen et du Havre’s catalogue. In doing so, we propose an organizational framework and pragmatical solutions for other contexts. We also argue this specific case to address more general issues for DH in the present day, i.e., how should we engage in making old digital critical editions sustainable? What are the sine qua non conditions to guarantee that an ‘ended’ project will nonetheless still be accessible in the future? And what does it tell us about the sustainability of digital editions more generally, and the role of secular infrastructures such as universities in that process?
Since the late 1990s, the issue of making digital humanities productions sustainable has become crucial in the field. On the one hand, specialists have emphasized the volatility of the Web technologies and the compelling difficulties in making a project sustainable in the long term (see the bibliography of the Endings project here). A convenient solution has been to develop guidelines and principles for ending a project elegantly (see Carlin et al. 2016 and the Endings project outcomes here). However, such solutions imply engaging with technical and organizational principles from the beginning of the digital edition project, e.g., in the choice of a given framework or a given architecture (see for example Smithies et al. 2019). It also requires to adopt specific practices, such as the uses of repositories, e.g., GIT, Zenodo, or Nakala. The reliance on dedicated and overly general infrastructures has also been criticized in the recent years (see for example Van Zundert 2012; Viglianti / del Rio Rande 2025)
What should we do, then, in the case an already ended project is to be preserved? Who should be responsible for making such a digital edition accessible? In what format (data only, data + interface, data + interface + documentation)? For how long? And what are the technical, organizational and legal conditions to achieve such a goal?
The case of Henein’s edition is enlightening because a) it had been an already finished and published edition when we faced the responsibility to preserve it; b) it originated from a period when no standards or rules governed digital editing; and c) no institutions were considered accountable for its sustainability at the moment we engaged in the project.
These are the main reasons why we consider this case to raise fundamental questions for the sustainability of digital critical editions more generally. Such a situation—being called upon to sustain and perpetuate a project initiated in the early 2000s—is, to our knowledge, not particularly common for a research team in the digital humanities today. Yet it is likely to become increasingly so.
We argue in the present paper that University Presses should be at the core of such movement, because they can rely on a secular organization with solid infrastructures such as IT services and libraries and because they have the institutional role of legitimizing knowledge production (Epron/ Vitali-Rosati 2018). Moreover, they are, because of their specific legal status, in a position to contract obligations towards the editors of a given critical edition, and to engage the University to ensure that such obligations are fulfilled.
The workflow we proposed thus consists of the following steps:
While it aligns with the customary process of print publication, this approach raised a series of fearsome and entangled questions.
First, to what extent are we producing a new edition in retro-converting the html hypertext edition to XML-TEI? Or is it just a form of digital ‘re-print’? Should we enhance the semantic annotation? Could we transform the edition architecture and the numerous relations between the numerous html pages? These technical and model-related issues will be discussed in details in the present paper, both because they are decisive in considering how digital critical edition could survive over the long-term, but also because they engage decisive organizational and legal issues.
The second series of issues we faced are related to the proposition we made for an ad hoc legitimation of the editorial work conducted by Eglal and Fekri Henein. Indeed, we argued that the Presses Universitaires de Rouen et du Havre should integrate the edition within their catalogue, with the aim to ensure the longevity of the edition. In doing so, Henein’s edition would be considered an editorial good similar to a printed book, and the University Press would be in the legal obligation to preserve it and to make it accessible. To do so, the edition of L’Astrée should nonetheless endure a reviewing process comparable to what exists for printed editions, a process that was not yet implemented at the Presses Universitaires de Rouen et du Havre. Since mid-2025, we have engaged in both creating a reviewing process dedicated to digital editions in this context, and worked on a suitable object to be reviewed and accepted.
In doing so, we also discovered that no clear legal framework existed at the Presses Universitaires de Rouen et du Havre to contract with editors about the sustainability of a given digital critical edition. How could the University legally institute the obligation to ensure the preservation and the accessibility of a digital edition? Will this concern only the data, or also the website interface? Under what conditions? For example, what are the material, equipment and human resources requirements imposed for University Presses to be able to achieve their obligations in this context? The present paper will present the result of a preliminary investigation on this subject conducted in France, with the aim to engage about such issues at a more global scale with the participants of the DH26 conference.
As we will show, our experience in Rouen with the case of L’Astrée digital critical edition raises a set of crucial issues for the sustainability and accessibility of digital editions at large, and digital critical editions in particular. While being limited to the specific context of France, it nonetheless opens important discussions about the function and necessary transformation of secular infrastructures such as University Presses. The solutions we are implementing will be, we hope, of interest for our colleagues in other countries, and we are convinced it would raise fertile debates and open new research collaborations on the subject.