Minimal computing offers a compelling formula for digital humanities projects. By forgoing complex, server-side operations, it promises to reduce dependencies on specialized software, sustained grant funding, and technical expertise that may not be available for all project teams or institutional contexts. These commitments — often articulated as using particular tools or methods, such as static-site generation, lightweight scripting, and minimal server infrastructure — have productively reshaped conversations about sustainability, labor, and technical equity in the field of digital humanities (Gil 2015; Sayers 2016; Gil and Ortega 2016; Risam 2019; Chan and Sayers 2022). At the same time, this tool-centered articulation of minimal computing has encouraged a widespread assumption: that relational databases sit outside its scope because they are too complex, too dependent on specialist labor, too fragile for long-term preservation, and too slow or resource-intensive for low-bandwidth contexts, particularly in the Global South. A narrow framing of minimal computing, however, risks conflating its values with a limited set of technologies rather than treating it as a design stance attentive to constraint, sufficiency, and tradeoff. This paper takes minimal computing’s core principles as its starting point, asking how they can be enacted when a project includes relational data.

Building on Risam and Gil’s (2022) reframing of minimal computing as a values-driven design philosophy that emphasizes the use of only “necessary and sufficient” tools rather than a fixed technical stack, our paper takes seriously the question of necessity. Risam and Gil (2022) propose a rubric for minimal computing grounded in four questions — What do we need? What do we have? What must we prioritize? What are we willing to give up? Their heuristic raises an underexplored possibility: what if a project needs a database? Drawing on our experience building the Pan-African Data Project, launching in 2026, we argue for “hybrid minimal computing” as an experimental design approach that combines elements of minimal and more complex infrastructures in order to support relational data while still prioritizing sustainability, accessibility in low-bandwidth environments, and reduced long-term infrastructure dependency.

The Pan-African Data Project is a data curation and digital mapping project designed to document the people, organizations, and events connected to Pan-Africanism from the late nineteenth century to the present. Pan-Africanism is a transnational political and intellectual movement that has brought together people on the African continent and in the diaspora to organize against colonialism and racism. Drawing on archival sources, including conference proceedings, newspapers, periodicals, declassified government documents, and scholarship in Black internationalism, the Pan-African Data Project curates a small but richly interpretive dataset comprising over 1,700 individuals, 45 events, and 43 organizations behind the movement. The project also incorporates data on Afro-Asian internationalisms in order to reflect the relational histories that linked anticolonial movements and postcolonial nation-building across the twentieth century.

In the historical record, Pan-Africanism survives in fragile and dispersed archival traces — unevenly preserved across continents, shaped by state surveillance, and deeply entangled with ongoing struggles for sovereignty, dignity, and liberation. The Pan-African Data Project reconstructs this history by bringing together participants and the events and organizations in which they were involved, alongside biographical, contextual, and geographic data. The public-facing website presents the data through interactive maps, tables, timelines, and information panels that allow users to investigate which individuals have relationships to particular places, events, and organizations, and, in many cases, to each other. In doing so, the project foregrounds exploratory, interpretive engagement with historical relationships and models the value of small, archival, humanities-driven datasets for postcolonial digital humanities scholarship that addresses omissions in the digital cultural record and surfaces histories marginalized by colonial and racial power (Risam 2019).

The central design challenge of the Pan-African Data Project was how to support this relational representation while remaining aligned with minimal computing principles to ensure the project would be accessible for users in the Global South, whose histories the project encompasses. Representing Pan-Africanism requires more than listing people, events, and organizations; it must maintain the relationships among them across space and time. A minimal computing model limited to page-based representations — the kind produced through static-site generators and flat-file workflows without explicit relational structure — would have reproduced the fragmentation of the historical record itself, scattering connected actors and moments across disconnected entries. At the same time, a continuously server-backed database would have introduced long-term risks incompatible with the project’s constraints, including ongoing hosting costs, maintenance burdens, institutional dependencies, and infrastructural precarity.

To address this tension, the project team organizes its technical workflow across two sequential stages: using a relational database during development and static public deployment for long-term access. The Pan-African Data Project adopts a hybrid minimal computing architecture that separates relational data modeling from public-facing presentation, treating each as a distinct phase of the project. During active research, the project uses a database-backed environment to maintain normalized records of people, organizations, events, and places — keeping “like with like,” avoiding duplication, and preserving explicit relationships among entities. This model is deliberately conventional rather than novel, drawing on well-established principles of database design precisely because those principles reliably support relational integrity.

The methodological intervention lies in when this database is used over the life of the project. Rather than binding relational data to a continuously running, server-dependent interface, the database is used only during development to maintain records and to generate maps, timelines, and tables, which can be published in static formats that do not rely on live database queries. In addition, “paths” (a storymapping feature) can be authored directly through a content management system, which also supports ingestion of custom data exports to create narratives with embedded data visualizations (“posts”); both can be published in static forms as well. Together, these approaches, developed for the Pan-African Data Project, offer an open-source, minimal computing alternative to cloud-dependent services such as ArcGIS StoryMaps and Observable.

By distinguishing between relational data needs during active research and static public access to visualizations and data-rich narrative materials, the team treats this boundary as a deliberate and productive constraint. The workflow enforces a conceptual distinction between data work and presentation work, which also enables different contributors (including students) to engage the project at the level of presentation without destabilizing the underlying data. Technically, the Pan-African Data Project draws on an open-source object-graph management framework that enables entities to be referenced consistently across multiple contexts, alongside a lightweight, git-backed content management system for producing paths and data essays.

Crucially, this approach enables the project’s public interface to be compiled and deployed as a fully static site. Static deployment — widely valued within minimal computing for sustainability, low-cost hosting, and accessibility in low-bandwidth contexts — is typically incompatible with relational databases and with cloud-hosted interactive platforms. The Pan-African Data Project addresses this tension by sequencing infrastructures over time: a full-stack, database-backed environment supports collaborative and intensive development during periods of active research and funding, while interactive maps, timelines, tables, and narrative materials generated from the relational core are delivered as static assets for long-term access. In this way, relational complexity is retained where it is necessary and released where it would otherwise impose ongoing infrastructural risk. While developed in response to the specific historical and ethical demands of Pan-African data, this approach offers a generalizable model and open-source software for digital humanities projects that require relational data but cannot commit to permanent, server-dependent infrastructures.

By treating relational databases as a scoped and temporary component within a minimal computing ethos — rather than as an all-or-nothing infrastructural commitment — the Pan-African Data Project demonstrates how relational rigor, interpretive depth, and long-term sustainability can coexist in digital humanities projects. Through this case study, we argue that a hybrid minimal computing offers a pragmatic, values-driven model for building relational digital humanities scholarship that can endure beyond periods of active funding while still remaining accessible to global audiences. Unlike ad hoc combinations of dynamic and static components, hybrid minimal computing formalizes infrastructure sequencing as a design principle, making explicit when and why relational systems are used — and when they are intentionally abandoned. Hybrid minimal computing does not resolve the tensions between minimal and maximal approaches; instead, it treats those tensions as design resources, allowing projects to combine relational complexity and infrastructural restraint in ways that are responsive to both historical materials and contemporary constraints.