A new open-source project published in the journal SoftwareX aims to change that. Developed by Yasiel Pérez Vera, the software presents an ecosystem of three interoperable plugins for Open Journal Systems, the widely used open-source platform that powers tens of thousands of scholarly journals worldwide. Rather than building yet another DOI workflow, the ecosystem implements two alternative persistent identifier schemes that carry no recurring registration costs: the Archival Resource Key, known as ARK, and the Persistent Uniform Resource Locator, or PURL. Both schemes offer decentralized, openly governed approaches to persistent identification. ARK identifiers support long-term stewardship through locally managed namespaces and resolver services, while PURLs provide stable redirection mechanisms that preserve identifier persistence independently of changes in the physical location of resources. The design philosophy is explicitly complementary rather than competitive: the ecosystem is not intended to replace DOI infrastructure, but to give journals that cannot sustain registration fees a flexible, sustainable alternative they can manage directly.

The software engineering story behind the project is as interesting as its economics. Conventional implementations of persistent identifiers in publishing platforms tend to tightly couple two very different responsibilities: generating identifiers and resolving them, meaning the act of translating an identifier into the current location of the resource it names. The new ecosystem deliberately separates these concerns. Identifier generation is handled by two plugins, one for ARKs and one for PURLs, each of which extends the native PubIdPlugin interface provided by Open Journal Systems. Identifier resolution, by contrast, is delegated to a third, independent component called the Resolver Gateway Plugin. This gateway receives incoming requests, identifies which identifier scheme is being used, retrieves the associated publication metadata from the platform’s database, and redirects the user to the appropriate resource. Because the gateway is generic, it operates independently of any specific identifier scheme, which means new identifier technologies can be added in the future simply by writing an additional publication identifier plugin that reuses the existing resolution infrastructure.

This architectural decision carries real practical consequences. Because all three components rely exclusively on the official plugin framework and extension mechanisms of Open Journal Systems, no modifications to the platform’s source code are required. Journals can install, update, or remove each component independently using the standard plugin management facilities, and the installation order of the plugins does not affect functionality. The ARK Plugin generates identifiers based on configurable Name Assigning Authority Numbers, object type prefixes, and local identifier patterns, storing them within the standard metadata model so they interact seamlessly with existing export, indexing, and publication services. Validation mechanisms prevent duplicate assignments and guarantee uniqueness within each journal namespace. The PURL Plugin follows the same principles, maintaining persistent identifiers separately from publication URLs so administrators can move content to new locations without invalidating existing citations. Crucially, both plugins can coexist within the same journal installation without conflict.

For journal administrators, the practical workflow is deliberately lightweight. To issue ARK identifiers, an institution first obtains a Name Assigning Authority Number from the ARK Alliance through an online registration form, a process that requires no fee. Once the namespace and resolver configuration are in place, the administrator configures the plugin through the standard settings interface, specifying the namespace, identifier pattern, target publication objects, and resolver URL. For PURLs, the administrator selects or operates a PURL resolution service and configures the corresponding namespace and generation policy. After that initial setup, everything happens automatically: whenever an article, issue, galley, or supplementary file is published through the normal editorial workflow, the appropriate plugin generates and validates an identifier, stores it in the publication metadata, and exposes it as an actionable link. Editors never need to touch identifier infrastructure again, and readers who click an identifier are transparently redirected to the current location of the content.

The validation of the software was conducted with unusual rigor for a publishing tool. The ecosystem was deployed on a dedicated server configured to replicate a typical institutional publishing infrastructure, hosting three scientific journals comprising approximately thirty published issues and 360 research articles. The functional validation protocol covered installation, configuration, automatic identifier generation, metadata persistence, and resolution through the gateway, with every test repeated thirty times under identical conditions. All twelve functional test cases passed consistently, and the validation processed 1,181 publication objects across every supported entity type, including journals, issues, articles, galleys, and supplementary files, without a single plugin failure or manual intervention. The tests also confirmed that all three plugins could operate simultaneously, demonstrating genuine interoperability rather than isolated functionality.

Performance measurements were equally encouraging. Using Apache JMeter 5.6.3 and the Linux time utility, the researchers measured the runtime overhead introduced by identifier generation and resolution across thirty independent executions of each operation. ARK generation took a mean of 15.8 milliseconds, PURL generation 14.7 milliseconds, identifier resolution 27.4 milliseconds, and complete HTTP redirections 31.2 milliseconds, with success rates of 100 percent for generation and 99.8 percent for resolution and redirection. These latencies are low enough that identifier management remains entirely transparent to editors and readers, an essential property for software that must sit inside a live editorial workflow without slowing it down. The deployment procedure itself was reproduced from clean installations of Open Journal Systems using only the official plugin installation mechanism, with no manual database changes, satisfying reproducibility expectations aligned with the FAIR principles for research software.

Perhaps the most striking evidence of the project’s relevance comes from its adoption. At the time of evaluation, 42 scientific journals across multiple countries had already deployed the ecosystem in production, with publicly verifiable, resolvable ARK and PURL identifiers confirmed through direct inspection of journal websites and live resolution requests. This is not a laboratory prototype but software supporting real editorial workflows on three continents. A comparison with conventional persistent identifier solutions for Open Journal Systems highlights the gap it fills: existing approaches typically depend on external registration agencies such as Crossref or DataCite, usually require recurring fees, offer limited support for multiple free identifier schemes, and lack an independent resolver component. The new ecosystem requires no external registration agency, imposes no recurring costs, supports self-hosted deployment, and remains fully extensible through the standard plugin framework, all while preserving compatibility with DOI-based solutions for journals that use both.

The broader significance extends beyond any single journal’s budget. Persistent identifiers are no longer valued merely for keeping citations alive; they increasingly serve as the connective tissue linking publications, datasets, software, researchers, institutions, and funders through interoperable, machine-actionable metadata, underpinning discovery, provenance tracking, reproducibility, and long-term preservation. When thousands of journals, particularly those in the Global South and the Diamond Open Access movement, operate outside the identifier ecosystem, the entire scholarly graph develops holes. By lowering the technical and economic barriers to participation, this open-source ecosystem, released under the GNU GPL v3 license with complete documentation and versioned releases in public repositories, offers a route toward a more complete and equitable identifier infrastructure. The author outlines future work including support for additional identifier schemes, improved interoperability with external scholarly infrastructures, automated testing and continuous integration pipelines, and validation across newer platform releases and larger publishing environments, positioning the software as a sustainable, community-extensible foundation for the next generation of scholarly communication.

Subject of Research: An open-source plugin ecosystem providing free ARK and PURL persistent identifiers for Open Journal Systems scholarly publishing

Article Title: An open-source ecosystem for implementing free persistent identifiers in open journal systems

Article References: Pérez Vera, Y. (2026). An open-source ecosystem for implementing free persistent identifiers in open journal systems. SoftwareX, 36, Article 103023. https://doi.org/10.1016/j.softx.2026.103023

Image Credits: AI Generated

DOI: Not provided

Keywords: persistent identifiers, Open Journal Systems, ARK identifiers, PURL, open access publishing, DOI alternatives, open-source software, scholarly communication, FAIR principles, resolver gateway, diamond open access, SoftwareX