1. Introduction

1.1. Scope

The CDL Guidelines for Digital Objects (CDL GDO, this document) provides specifications for all new digital objects prepared by institutions for submission to CDL for access and preservation services. They are not intended to cover all of the administrative, operational, and technical issues surrounding the creation of digital object collections.

The guidelines seek to support the following objectives:

• Ensure a basic level of uniformity in the structure and encoding of non-licensed digital content managed by the CDL
• Advance interoperability among digital content from diverse institutions
• Promote efficient ingest procedures
• Support the orderly management of digital content
• Facilitate access to digital content by users
• Minimize costs

These guidelines do not set requirements for digital materials submitted to or collected by the CDL through other means:

• Metadata exposed to CDL harvesting systems via the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH)
• Metadata targeted by federated search systems
• Web-crawled resources

In addition, these guidelines do not address requirements for collections delivered to the CDL/University of California Office of Scholarly Communication’s (OSC) eScholarship Repository or for the UC Image Service. Institutions interested in submitting content to these repositories should consult the projects' websites.

1.2. Service Levels

Digital materials of ever-increasing variety and complexity are seen to be worth collecting and preserving by memory organizations such as libraries, archives, and museums. Materials include objects converted into digital form from existing collections of manuscripts, maps, visual images, and sound files, as well as born-digital materials such as web sites, videos, and data sets. Submitted objects consist of metadata, a set of content files, and something called a METS digital wrapper file.

In order to create coherent and cost-effective services for such diverse collections, the CDL and other digital libraries sometimes require certain common digital object features that offer strategic points of leverage. This is a delicate undertaking, as it tends to involve a reduction in diversity that implies a loss of information, and every imposed requirement incurs the risk of rejecting valuable materials that fail to meet it. Simply meeting requirements is often hard because funding is unavailable or the original producer of the digital objects cannot be reached.

To mitigate these difficulties, the CDL adopts "sliding scale" guidelines: the more points at which a digital object can be made to conform, the more preservation and access services can be provided for it. The CDL GDO sets forth minimum submission requirements for digital objects submitted to the CDL.

At the lower end of the scale, given no information about the structure and semantics of a set of files comprising an object, the level of preservation that we can promise is limited to bit-level preservation and identifier-based retrieval -- the bits of a digital object that you submit will be the same bits that you are able to retrieve, and the only access is by known identifier or by any internal data that happens to be discoverable and indexable (e.g., content files in the form of text). In short, no metadata is required. However, providing metadata may allow access to additional preservation services while also enhancing value for future users.

At the mid-level of the scale, given a small (or "kernel") set of structured metadata encoded in a METS wrapper, the CDL will be better equipped to manage the objects and provide preservation and kernel metadata-based retrieval services. (For more information about kernel metadata, see the Dublin Core website).

Moving up the scale, generally speaking, the more metadata encoded in a METS wrapper that you supply, the better we will be able to provide you with preservation and access services. Our systems may not be able to take advantage of every distinct metadata element that you supply, but the ability to act on any element may be developed over time as our systems evolve. The higher end of the scale includes the ability to customize the formatting and grouping of collection objects, which depends on a combination of XSL style sheets and your provision of metadata elements that our systems can recognize.

The CDL GDO specifies requirements for two primary levels of services offered by the CDL:

• Basic Service Level: sufficient for the ingest of digital objects into the UC Libraries Digital Preservation Repository (DPR), this level is designed to support the orderly management of objects in the DPR, hence our ability to provide at least bit-level preservation without turning away valuable materials. It currently does not support the presentation of digital assets via CDL websites.  This service level does not require any metadata, but strongly encourages kernel metadata. A range of content file formats is supported at this level.
• Enhanced Service Level: includes the presentation of digital assets via CDL websites. It is also sufficient for increased preservation services in the DPR. This level is a detailed extension of the Basic Service Level digital object specification, and therefore prescribes for additional metadata encoding. Particular content file formats are supported at this level.

1.3. Terminology

For an explanation of general terms used throughout these guidelines, see the CDL Glossary. For an explanation of concepts and terms pertaining to metadata in particular, consult the RLG Cultural Materials Descriptive Metadata Guidelines.

1.4. How to Use These Guidelines

Consult the appropriate section of the guidelines, based on the level of CDL service that your institution is interested in utilizing:

• Basic Service Level: consult Section 2 only
• Enhanced Service Level: consult Section 3 only