Go to National Library of Australia Home PageNational Library of Australia Online

SEARCH:

APPENDIX 4
Managing Web Resources for Persistent Access

The success of a distributed information system such as the World Wide Web (www) for research depends on the long-term consistency of the interlinking between online resources and the persistence of the links that are provided in the catalogues, indexes and listings which constitute resource discovery services.

Links between resources can range from casual links inserted in a document because it is possible to do so, (e.g. a link to the site of an organisation because the name of that organisation is mentioned), to the formal citation of references, source documents, research results or to other related information that is an integral part of the resource. The category of link that represents formal citation is the most significant for the integrity of the intellectual heritage and it is especially important that these links continue to identify and provide access to the resources.

The persistence of links between resources and of links in resource discovery services is essential to ensure long-term public access to web based materials and is therefore an important aspect of the archiving and preservation strategies adopted for these resources. For information on how to manage web resources to assist later preservation action, see the companion document Safeguarding Australia’s web resources: guidelines for creators and publishers

Creators, publishers, libraries and other information repositories all have a vital role to play in ensuring web based information resources can be located and accessed for use in the current environment and into the future. These guidelines are intended to assist those responsible for the management of published online materials to make sure links made to those resources continue to work.

1. Why links fail to work

Broken links are very common on the web and few people will not have come across an error message saying that the resource is no longer available.

Among the most common reasons why links on the web fail to work are:

  • The resource may have been removed because it is no longer current or no longer perceived to be of value by its owner – ie. it “disappears”
  • The resource may have been moved as a result of a file restructure, a change of domain name or a change of ownership and it has therefore acquired a different URL

Documents which “disappear”

A document may be removed from a website for many different reasons, including.

  • it no longer reflects current policy
  • the information in it has been superseded
  • a hard copy exists for preservation
  • there is an official copy in the electronic records management system
  • the publication is no longer of commercial value
  • “no one will want this any more”
  • there is no place for it on the redesigned website

When considering removing a document from a website bear in mind that

  • although you may no longer appreciate the value of your older documents, they may be valuable to others for continuing reference, research or may be of historical interest
  • keeping the documents available can add to the depth of content on your site
  • there could be links to the document in other documents and in catalogues and indexes
  • broken links to documents on your site reduce its credibility

Documents which are moved

When considering moving documents around a site, bear in mind that

  • unless measures are taken to point to the new location, all existing links to the document are broken once it is moved.
  • there could be links to the document in other documents and in catalogues and indexes

2. Managing resources to ensure persistence of citation

The maintenance of links to resources on the web is essentially a task of good management. It involves planning ahead and taking deliberate steps to avoid the unnecessary removal or movement of documents on the site. The following guidelines will assist with developing a data policy for your site that should be used to manage resources for persistent access.

Determine document characteristics

Take a look at the types of document you have on your site, or on your organisation’s site, and see what categories they fall into. Resource categories which need persistent access include

  • information which is likely to be of long term value
  • resources which you know will be linked to by other resources
  • resources that are bookmarked by many people

There will also be transient or temporary information, which can be removed.

The information which needs to be managed for persistent access will generally fall into two types

  • static resources
  • dynamic resources

Static resources include items which have a definitive version and will not be altered, e.g. a published policy document, an annual report, a conference paper, all of which reflect the point of time at which they were created.

Dynamic resources are materials which undergo change, either incremental change by addition of content or change by modification, deletion and insertion of content. They can be either “archival” or ephemeral.

The “archival” category includes material that you wish to preserve and provide persistent access to. The decision about what level of persistence of access to provide will be a matter of judgement and will depend on the nature of the material. E.g. for an electronic journal which operates in an incremental fashion, it may be necessary to provide consistent access to both the “home page” and to individual issues and articles for citation purposes; for a publication which may be modified by change, addition and deletion, you may wish to consider archiving and providing access to successive versions so that little information is lost and the evolving nature of the publication is documented.

The ephemeral category includes “What’s new”, “What’s on”, opening hours, contact information, current publications, etc. These are not likely to be cited in other resources, but are frequently bookmarked for regular use and users want the latest version. Old versions are generally of no particular interest and this category needs persistence of citation only for the current version.

Once you have determined what categories and type of material you have, decide how you will ensure that citations to the material in the form of links remain active and enable the user following the link to access the resource.

Maintain active links

There are many ways to manage the material so that citations remain active. They fall into three broad categories:

  • Use a persistent identifier system
  • Organise the website to reduce the need to move material around.
  • Keep older material accessible even if it is no longer on the main pages of the website.

Use a persistent identifier system

Most resources currently on the www are described using a Uniform Resource Locator (URL) and the URL as generally used describes a resource in terms of its current location, i.e. it describes the domain name of the server, directory path and the document name of the resource.

The location based nature of the URL is its major weakness for document naming, since the identifier will change if the material is moved either to a different server or a different place in the file structure of the same server. When this happens, all links to this resource embedded in other documents or databases will be broken.

A persistent identifier is a name for a resource which will remain the same regardless of where the document is located. Use of a persistent identifier will ensure that when a document is moved, or its ownership changes, the links to it will remain actionable.

It is important to note that a persistent identifier system will only be effective if it is maintained. When resources are moved, the current location must be associated with the persistent identifier in whatever system is chosen. This is usually achieved by a resolver database. There is no system which does not require such management.

Your choice strategies to achieve persistent identification will depend on your circumstances. Some options are:

Use redirects to guide users to the current location of the resource

As a minimum, the standard redirect capabilities of the web server can be used to redirect requests for resources using a persistent identifier to the actual location of the document or resource. This method is hard to manage on a large site and records of redirects in place need to be kept.

Use a resolver database

A resolver database can be used to map the name to a current location. Discuss this option with your webmaster or ISP, as there may be other people who have material on the site who may also wish to use such a facility. One option for achieving this is to mount a PURL server on the site. The PURL server software is made available for downloading from OCLC’s PURL site http://purl.oclc.org/.

Use a persistent identifier service offered by another organisation

There are a number of other persistent identifier systems designed for use on the Internet. You may wish to investigate the use of one of these, e.g. DOI (Digital Object Identifier) http://www.doi.org/ or a Handle system http://www.handle.net. OCLC offers a PURL service to organisations that do not wish to mount a PURL Server on their own site. Users can register and apply online for a PURL and maintain their PURL(s) by updating the location of the resource online if it is moved. http://purl.oclc.org/

Persistent identifier services and the National Library

The National Library is evaluating options for offering a persistent identifier service.

Organise the website to reduce the need to move material around

Even if a persistent identifier system is not used, a website can be organised and managed to reduce the need to move material around.

Organisation of the site

The directory structure in a site can be structured and named in a way which reduces the likelihood of change being necessary. Guidelines for consideration include:

  • Avoid long directory paths and try and use logical names rather than addresses in a directory structure so that file names do not need the directory path to make them unique and the temptation to reorganise the drectory structure to accommodate new resources is reduced
  • Avoid putting publicly accessible resources in private directories so that when the owner of the directory moves from the organisation or hands over responsibility for the resource, the URL will not have to change
  • Avoid the use of transitory organisational names in the directory structure to reduce the temptation to change the directory structure and hence the URLs when the organisation ceases to exist or changes its name
  • Consider using a date as far up the directory structure as possible for parts of the website. This can help to organise the website in a way which may assist in the management of static resources or versions of dynamic resources by indicating the date of creation and hence the standards used to create them
  • If a query string is used for a known item search, don’t encode system related data in the field, as the system will be replaced in time

Re-organising the website

You do not have to re-organise the directory structure of a website to revamp the site for a “new look”. The navigation and presentation features of a website are completely separate from the underlying organising structure of the site. The site can be given a totally new look by redesigning the graphics, presentation and navigation without moving any of the underlying files on which it is built.

Keep older material accessible

Older documents which have been superseded need not be removed from a website just because they no longer reflect current policy, or because their factual content is dated, or simply because they are no longer “flavour of the month”. Older content is valuable to researchers and policy developers and can be of interest to others. As stated earlier, there will be links to the documents from other sources which will be broken if the document is removed.

To keep older documents, you do not necessarily have to retain them on the main pages of the site. They can be:

  • accessed through a logical archive
  • moved to a physical archive on the site
  • lodged with an archival organisation such as a library for preservation

Access through a logical archive.

When documents are no longer presented on the main pages of a site, they can be made available through a logical archive. The documents themselves are not moved, but access to them is through some form of “archive listing” on the site. Because they have not moved, the documents remain accessible via the links that have already been made to them in other resources.

Moving to a physical archive on the site

Documents may also be moved physically to a section of the site designated as an archive for management purposes and accessed via an archive listing or a search engine run over the archive. For long-term preservation, this is a safer way to proceed, since it simplifies the management process. If this method is chosen, a pointer or some other form of mapping will be needed to direct users to the new location.

Ideally, items would be identified for long-term preservation at the time of creation and would be located in such an archive when initially mounted on the server.

Lodging with an archival institution

If you cannot maintain access to the resources on your own site, then consider asking another body such as a library or an archive to take over management of the document for you. The National Library, National Archives of Australia, your local State library or archive or university library may be willing to preserve it if it is considered to be within the scope of their archiving policies. Again, if this is done, a pointer to the new location will need to be maintained.

If an older document is “archived” using any of the above methods, you may wish to display a notice with the resource stating that it no longer represents current policy, or advising that the content has been superseded and directing the user to more up to date information.

For further information on steps to take which will assist the preservation process for resources published on the web, see the companion document “Safeguarding Australia’s web resources: guidelines for creators and publishers” http://www.nla.gov.au/guidelines/2000/webresources.html>

Contact

For more information or assistance with any matter raised in these guidelines, please contact:

Manager
Electronic Unit
National library of Australia
Canberra ACT 2600

Email: elecunit@nla.gov.au
Fax; 02 6273 4322
Phone: 02 62621140

March 2001

[Note: this document has been published as a brochure as a companion brochure to Safeguarding Australia’s web resources: guidelines for creators and publishers]

Return to Table of Contents | Proceed to Appendix 5
About This Site | Copyright | Privacy | Accessibility | Site Map | Site Search | Content A-Z | Contact Us