A consultative document on the content of a “DLM-Message to the ICT Industry”
Beratungsdokument des DLM-Forums für Archivstandards zum Inhalt einer Botschaft an die Industrie.
Executive Summary
The users of information management software are not a homogenous group. While the software industry may have a broad recognition of records management principles and the need for their support by software products, it should be aware that software products developed for specific areas or particular user groups may not have universal application or be portable to all client situations. However, despite the heterogeneity of the information management software market, certain key issues have been identified as impacting on most users. These include requirements for software that supports a standard functionality set, the re-use, interchange and long-term storage of information, open standards and specifications, metadata standards, non-English languages, records containing graphic video, and audio content, interfaces with public key infrastructures including electronic signatures and message digests, mark-up languages, etc. In addition, there is a need for industry codes of practice to address the areas of concern referred to in this document. While it is the responsibility of information professionals to state their business needs, it is, of course, for the software industry to provide solutions. But it is not sufficient for information professionals to state their requirements in broad terms and to expect industry to react. Broad business needs should be defined in greater detail in calls for tender, and information professionals must ensure that their specific requirements are included in their organisations’ purchasing strategies. For it is only by demonstrating that a genuine market demand exists that the software industry will respond and produce the systems and services required to support the management of electronic information.
A. Who we are:
Industry should realise there is no single user community for information management software. Typically, users include:
| | | |
| | · | Information Handling Professionals, (i.e. decision makers in administration and in enterprise, archivists, information managers, ICT-professionals, etc.); |
| | | |
| | · | Information Users (i.e. business process managers, knowledge workers in, for example, the pharmaceutical industry, the civil service, the educational sector, the legal profession, insurance, banking, healthcare, media, etc.); and |
| | | |
| | · | Clients (i.e. member of the public and businesses that wish to access information holdings maintained by others and exchange information with them). |
All these groups have different business and technical needs and all have different levels of exposure to, and experience of, information technology. Therefore, the software industry should be aware that a software product designed with the needs of one user community in mind may not be directly portable to another, even where it supports records management principles. Ideally, information handling software products should be able to address a variety of needs and markets, and have the flexibility to be adapted to a range of records management practices.
B. What is required:
While any heterogeneous group will have multiple and varied requirements there is a common core of key issues that must be addressed satisfactorily by industry.
These include:
1. Functional Requirements
It would be desirable for the software industry to state clearly the functionality of information management software and its envisaged target audience. Where accepted records management standards and methodologies exist industry should state if their products comply with such practices. In this area, a particular requirement is to know if a product supports the functionality to manage an electronic record throughout its whole life cycle.
2. Longevity and Disposal
The short life of electronic records, due to rapid technological change, is a particular cause for concern. We must recognise that the electronic records of today are the archives of tomorrow. Therefore, they must be accessible over long periods of time. While it is absolutely necessary for archivists to define and quantify the requirements for the long term preservation of records, it is crucial that industry meets stated requirements in this area, and, where requested, provide guarantees on the longevity of persistent information and storage media.
Where such guarantees cannot be given it is essential that industry provide clear upgrade paths that maintain both content and presentation when information is moved from one version of a software product to another version of the same product, or between similar software products from the same manufacturer. Likewise, it is also essential that industry provide mechanisms to migrate both content and presentation to software products different from those used to create the original records. For this to happen, industry should specify and agree interchange formats among software providers.
It should be noted that this problem occurs, not only for textual information, but for all kinds of digital information. For example: numeric data, statistical information held in databases, spreadsheets used for decision making, images, etc.Industry should also be aware, that, in certain situations, not only must the records be preserved but also auditable details of the computer systems that produced them. Besides guaranteeing the long-term preservation of electronic records, it is equally important, both because of the costs involved and the usability of the systems, that records management systems include mechanisms to authorise the destruction of information that has only temporary value. This should occur after the expiration of predetermined retention periods.
3. The Re-use of Information
Neither should we forget the re-use value of information. The information held in both public and private sector archives constitutes an enormous wealth of content that could be exploited. These policy issues were addressed in the European Commission’s Green Paper on Public Sector Information in the Information Society – ‘Public Sector Information: A Key Resource for Europe’. However, unless the capability for the re-use of electronic records is built into the management systems that maintain and store this information, it will be, in effect, locked in local repositories, and the benefits from re-use, matching, exploitation, etc. will not be realised. Furthermore, it will impact adversely on the ability of our institutions to deliver timely accurate information to the citizen and to enterprise.
This is not purely an issue for archivists. In today’s organisation, the electronic document that was created in a word processing package or perhaps downloaded from a web site, may need to move seamlessly to the organisation’s workflow system, and then to an e-commerce system. It may leave the organisation via the e-mail system for processing elsewhere, and return for further processing in the accounting system, finally to be stored in a long-term repository. In each of these operations new data and additional metadata will be added to the document. The economic benefits of such information re-use are clear, but to achieve them further work is required to agree standards, interchange formats, APIs, etc.
4. Metadata
It is essential that information management software supports internationally recognised metadata standards There is considerable debate taking place, at present, among information professionals on the requirements for metadata and on defining metadata standards. Clearly, it would be desirable for industry to follow and participate in this debate and to support whatever recommendations are made for European and international metadata standards. Furthermore, there is a lack of work on metadata standards to describe technical computer environments and industry should contribute to this.
5. Open Standards and Specifications
The requirement for software products to be written to open standards and specifications has been stated so often before. However, it must be re-emphasised here. We know that industry will continue to promote its own proprietary formats for adding value to record creation/maintenance. In addition, users may well prefer to use such formats for reasons relating to ease of use, additional functionality, improved productivity, etc. However, software that exclusively uses proprietary formats causes particular problems in the areas of content interoperability, information interchange, re-use, longevity, migration, etc.
The need for agreed open interchange standards and formats and their equivalents for persistent storage cannot be stated too many times. There is a general recognition that standards-conformant products will make perhaps the major contribution in overcoming the problems in this area.
6. Linguistic Diversity and Multicultural
Issues
While document preparation software can, in general, support non-English languages, serious problems still occur where documents containing characters with diacritical marks, ligatures, etc. are transferred between software products. Again, there is a need for industry to agree interchange standards, base encoding standards, etc. for non-English character sets, and to inform customers how to configure their products easily to allow such interchanges to happen.
This issue is not solely related to the support for non-English character sets. Due consideration should be given to the linguistic diversity in Europe. Where a country has several official languages, software may be required to support multiple character sets in the same document. This may also be a requirement for firms operating internationally. This is not purely for aesthetic reasons. Frequently, countries and regions have legal requirements concerning the depiction of information in culturally correct language and presentation formats. However, the issue goes beyond simple support for multiple character sets. To be truly multicultural, software should also support different national presentation contexts (i.e. alphabetic sort sequences, conventions for positive and negative answers, address formats, etc.). Increasingly, these are standardised in national cultural registries, thereby providing a specification to which industry can incorporate multicultural support.
7. Multimedia
Increasingly, documents containing graphics, images, video, and audio, etc. are prepared using products that usually support proprietary formats only. Again, there is a need for industry to provide minimum agreed interchange standards to allow non-textual content to be migrated to other software products, and to be preserved in the long term. However, this issue is complicated by the rapid developments in technology where 3-D and virtual reality worlds are being introduced. These new formats, again, are largely proprietary, but systems are being developed that use them, for example, to record cultural content. Furthermore, the integration of all these media types is another key issue.
8. Public Key Infrastructures
Another major requirement will be for records management software to attest to the integrity of electronic documents, to authenticate their authorship, and to provide confidentiality for sensitive information when stored or transmitted over insecure channels. While standards for message digests, electronic signatures, etc. exist, and their legal recognition is now legislated for at both European and national levels, the same problems of interoperability and longevity occur again. This is a new and additional challenge for the software industry. Not only must content and metadata be interoperable and persistent, but now the message digests and electronic signatures associated with electronic records must also be readable by different systems, preserved for long periods of time, and verifiable at later dates.
9. Support for Mark-up languages
Apart from difficulties relating to the retention of content and presentation that arise when information is moved between IT systems, recipient systems cannot automatically process information unless programmed to identify and deal with specific information types. The use of mark-up languages, conforming to the XML standard , to make information self-describing should contribute to the solution of these problems. It would be desirable for industry, working in close co-operation with user groups, to support internationally recognised user group-specific mark-up tag sets (e.g. tags relating to clinical information) to describe the information being exchanged, and for software products to support agreed XML tag sets. However, for reasons relating to longevity and re-use of information it is the ‘openness’ of the information object, whether document, file, or database, rather than the software that produced or converted it, that is the crucial issue.
10. Industry Codes of Practice
It would be desirable if industry were to develop a code of practice to address issues such as:
Liability
Information professionals (e.g. archivists) may need to access records created by software that was originally licensed to the organisation that created the records. Liability issues might well arise where a third party, such as an external archivist, uses such software to access the records. It would be desirable if industry were to agree a convention or code or practice to allow information professionals to use freely, for legitimate research and archival purposes, software for which they do not have a licence.
Code Escrow
It may be impossible to read records created by software that is obsolete, or where the software providers have gone out of business. It would be desirable if industry were to agree code escrow practices whereby obsolete software could be recovered for use by information professionals. This is particularly important where bespoke software is developed for a particular organisation or a particular function.
C. How should it be provided?
The above is a list of the minimum requirements for software to manage electronic information. Many information professionals may feel that it ignores their particular speciality, while others may disagree with aspects of its content. However, unless the business requirements of information professionals are clearly stated industry will not be in a position to respond or will respond with inappropriate offers.It is not the function of information professionals to provide solutions – this is clearly the responsibility of industry. But for the software industry to react, business needs must be spelled out in greater detail and included in calls for tender and procurement exercises. Furthermore, to show that a genuine market exists information professionals must ensure that their business and technical requirements are included in their organisation’s purchasing strategies. For it is only when a genuine demand is demonstrated that the software industry will react by supplying appropriate open and standards-based information management software and services.
Metadata