String of PURLs: Frugal Migration and Maintenance of Persistent Identifiers

Tracking #: 603-1583


Responsible editor: 

Michel Dumontier

Submission Type: 

Resource Paper


FAIR data requires unique and persistent identifiers. Persistent Uniform Resource Locators (PURLs) are one common solution, introducing a mapping layer from the permanent identifier to a target URL that can change over time. Maintaining a PURL system requires long-term commitment and resources, and this can present a challenge for open projects that rely heavily on volunteers and donated resources. When the PURL system used by the Open Biological and Biomedical Ontologies (OBO) community suffered major technical problems in 2015, OBO developers had to migrate quickly to a new system. We describe that migration, the new OBO PURL system that we built, and the key factors behind our design. The OBO PURL system is low-cost and low-maintenance, built on well-established open source software, customized to the needs of the OBO community, and shows how key FAIR principles can be supported on a tight budget.


Previous Version: 


  • Reviewed

Special issue (if applicable): 

Special Issue on FAIR Data, Systems and Analysis

Data repository URLs: 


Date of Submission: 

Wednesday, August 14, 2019

Date of Decision: 

Tuesday, August 27, 2019

Nanopublication URLs:



Solicited Reviews:


Meta-Review by Editor

We are pleased to inform you that your paper has been accepted for publication, under the condition that you address the remaining issues: 

* That the contributions of this paper are clearly stated and made relevant beyond the direct utility of the work for the OBO community. 

* section: controlled x 2: we controlled controlled the upper

* previously reported, still appears: the system

Michel Dumontier (

Second response to reviewers

We would once again like to thank the reviewers for their helpful comments.

Reviewer 2 has clarified that he was asked to evaluate "does it provide sound evidence of its (potential for) reuse?" and that we have not provided sufficient evidence. This was a misunderstanding on our part of the review criteria, which we believe have been revised to include more information about the new "resource paper" category since we originally submitted our manuscript. In any case, with our misunderstanding resolved, we are happy to address this review criterion.

We have updated our GitHub repository to make it clear that our code is distributed under a BSD3 open source license.

We have added a paragraph to the "Future Work" section explaining that our code is free for other communities to reuse under an open source license, and that by adapting our YAML configuration format to suit their own needs they run their own PURL server to can gain the same benefits that we have over the system.

We have also made the two typos which were pointed out to us.