Guidelines for Web-Disseminated Records
The University of Michigan conveys much information and conducts many of its administrative activities via the web. The following guidelines are divided into two categories: design and preservation. The design section covers emerging best practices to aid in making the website usable and understandable in the future. 1 While it is certain that technologies will continue to change, the preservation or the "archiving" process will be more efficient when there is consistency and standardization embedded in the creation of the original web file. The preservation section includes guidelines on two possible approaches for the capture and maintenance of web-based records.
Each unit should identify and evaluate the types of records created and maintained on its website. Procedures should be established to capture records with administrative and long-term historical value into a secure recordkeeping system.
Best Practice Example [Word document]
The following design elements will better insure that a web page maintain the content and "look and feel" in the future as it was originally presented.
- Use standard HTML markup when tagging university webpages. HTML versions 3.2, 4.0, and 4.01 are recommended and are backward compatible with HTML 1.0 and 2.0. For further information on HTML standards see The National Center for Supercomputing Applications (NCSA) HTML Primer. Thinking of changes in web technology, the best practice would actually be to create HTML pages that are XML compliant.
- Include the name of the unit, date the information was last changed and contact information on each page. A good place for this information is the footer section of each page.
- In order to preserve the functionality of the original webpage, minimize the dependence on proprietary software applications by using non-proprietary file formats. These formats include: Graphics TIFF and JPEG; Video MPEG.
- Note that while Adobe Acrobat Portable Document Format (PDF) is a proprietary format, the PDF format seems to be moving towards a de facto standard for digital document distribution especially effective dated materials such as brochures, bulletins, reports, and other publications. Like with other technology dependent documents, you will want to track the software version used in the creation of PDF documents.
- Note that the functionality provided by Java script, plug-ins, third-party search engines, and databases is problematic to preserve overtime.
- In order to preserve the original navigation within the website, build hypertext links using relative Uniform Resource Location (URL) path addresses rather than absolute path addresses.
- Use Dublin Core Metadata Elements to aid search and retrieval, and provide documentation about the website. It is recommended that each unit include some description about the website within the website's main webpage. The description is embedded within the tags of the HTML file and is not visible to the user. Click here for an example of a metadata template using Dublin Core elements.
- Lastly, because a webpage can constantly change--think about how changes will be managed. One model is to systematically upload changes at a specific point in the term. Another model would be to create an audit trail. Suggested elements for a audit trail or history log are noted in the section on Website Preservation Guidelines for Active Management of Website Content
1. Emerging best practices sources used in the compilation of the web-based records section of this manual include Charles Dollar, Archival Preservation of Smithsonian Web Resources: Strategies, Principles, and Best Practices, Report (July 2001) and Charles R. McClure and J. Timothy Sprehe, Analysis and Development of Model Quality Guidelines for Electronic Records Management on State and Federal Websites, Research Study funded by National Historic Publications, Final Report (January 1998).