Telstar

5.3.1 Persistent Links

When providing links to online resources it is clearly desirable that the links will work over long periods of time. This toolkit does not attempt to cover all aspects of persistent identifiers, which are covered elsewhere; a good starting point is Emma Tonkin’s 2008 article on the topic in Ariadne.

In 1998 Tim Berners-Lee wrote “Cool URIs don’t change”. This document lays out the reasons why those responsible for maintaining URIs should ensure that they are persistent over time. When recording a reference to an online resource, if a ‘cool URI’ is available, it may well be appropriate to use it. However, judging if a URI is ‘cool’ or not is not easy. This essentially comes down to the question of how much you trust the provider of the URI will maintain the URI, the content it identifies, and it’s ability to do so over a long period of time. This is not simply a question of the intentions of the provider, but also relates to questions such as the long term viability of the provider as an entity (e.g. for commercial providers, what is the likelihood they will become insolvent, and in this case the likelihood of the links being maintained by another party).

Many providers of online scholarly resources now provide so called ‘persistent links’ which are intended to provide ‘cool URIs’ for their resources. It is worth noting that the ‘persistent link’ may not be the one that appears in the browser address bar when you browse to the page, but provided separately with other metadata about the document. Once again, how persistent you believe these links to be is a matter of how far you believe the publisher or provider will maintain the link (and be in a position to do so).

5.3.1.1 PURLs

While the term ‘persistent link’ has no formal definition, the term ‘Persistent URL’ (PURL) has a specific meaning. PURLs are:

“Web addresses that act as permanent identifiers in the face of a dynamic and changing Web infrastructure. Instead of resolving directly to Web resources, PURLs provide a level of indirection that allows the underlying Web addresses of resources to change over time without negatively affecting systems that depend on them.” (http://purl.org/docs/index.html)

The best known PURL server is managed by OCLC, although anyone can use the service to create PURLs, but PURL servers can be implemented by anyone using the PURLZ server software developed by OCLC and Zepheira, or using any software that can provide similar functionality.

As with any ‘persistent link’ the persistence of a PURL depends on the persistence of the service that resolves the PURL to a web resource. Implementing a local PURL service may increase local control over the service, but institutions should be realistic about their ability to maintain such services over long periods of time and the commitment this requires.

5.3.1.2 DOIs

In addition to persistent links supplied by individual providers, many scholarly resources are now being assigned ‘Digital Object Identifiers’ (DOIs). DOIs are intended to be persistent identifiers (rather than persistent links), but can generally be turned into links to the online resource (a process known as ‘resolving’) by appending the DOI to the address of a resolver – the best known being http://dx.doi.org/.

As an example the DOI “10.1037/0003-066X.59.1.29” identifies the article:

Ray, O. (2004) ‘How the Mind Hurts and Heals the Body’, American Psychologist, Vol. 59, No. 1. pp.29-40.

Which means that you can link to the online version of the article with the URI http://dx.doi.org/10.1037/0003-066X.59.1.29

References in scholarly works now commonly include a DOI when one is available. In these cases it is possible to create a link to an online version of the resource using a DOI resolver. However, the DOI will usually resolve to a specific version of the resource online, usually the one provided directly by the publisher, and if the reader does not have access to this version, they will be unable to view the resource via this link, even if they have access to the resource through an alternative route. This is the ‘appropriate copy’ problem in action.

5.3.1.3 OpenURLs

An alternative approach to linking to scholarly resources online is to use ‘OpenURLs’ (or more formally The OpenURL framework standard ANSI/NISO Z39.88). The OpenURL framework enables the creation of applications that transfer packages of information over a network. The only significant implementation of the standard is to transfer metadata related to bibliographic resources.

OpenURL has seen widespread adoption by University libraries in combination with ‘OpenURL resolver’ software. Resolver software typically uses the metadata available from an OpenURL (transported over http) to provide a link to the ‘appropriate copy’ based on the library’s subscription information.

As OpenURLs are used to transport metadata, they can include other appropriate identifiers such as DOIs or ISBNs (International Standard Book Numbers) either of which can be used to form links to online versions of the resource – if the resolver is configured to do this.

OpenURLs should not be regarded as ‘persistent links’, as their contents (and therefore the link) can change over time (if the source metadata is updated for example), and the outcome of following an OpenURL link can change depending on the configuration of the resolver software being used. However, they do provide a consistent mechanism for linking to a copy of the resource via a library or institutional subscription. As OpenURLs are constructed from the reference metadata they can be formed ‘on-the-fly’ from the reference, if this is desired.