Skip Navigation / Jump to Content


Answered questions

Return to the ErpaAdvisory index page | View list of unanswered questions

There are currently 21 answered questions on ErpaAdvisory:

Questions 11 to 15 shown below.

Previous | Next

Submitted by swissed on 24 April 2003 at 7:10

When reading about digital preservation methods I sometimes come across recommendations to print electronic files out to paper. Is this a serious proposition, and in what cases might I consider it?

Answered by swissed on 24 April 2003 at 7:12


This question is about the digital preservation method commonly called “print to paper” that involves printing out relevant digital documents and preserving the paper copy instead of the digital original.


While printing out the digital documents and archiving the paper printouts is discussed and employed in some circumstances as a preservation strategy, this cannot properly be considered digital preservation, since it implies a transfer of digital documents to a non-digital state. The method faces serious drawbacks. A record’s authenticity depends, among other things, on its digital or analogue form. Changing this state implies possible loss of informational and/or evidential value, might provoke uncertainties about which one is the original or the valid version and raises concerns about how to deal with other, concurrent versions. In fact, recent legislation and court case have tended to favour records presented in the format they have originally been created in, and therefore in certain cases, digital records must be preserved digitally (see below). Of course, if the documents concerned are not records, retention requirements will be different, and printing to paper may be a more viable option. (The benefits of this method, namely reducing the digital preservation problem to a better understood one of paper or microfilm preservation, can be taken into account.)

In addition to this, there are some evident prerequisites for applying this method. Printing a document is only possible for comparatively simple, static, and flat documents like word processing files, image files, or simple data tables (but not for databases, multimedia documents, and any other more complex document type), and if the digital functionality is no longer needed.

_Legal Issues_
Legal requirements for the management of records exist in both substantive and procedural law. Substantive law dictates how organisations must keep records (taxation, accounting, and contractual law). Evidence law and rules of admissibility dictate that original is to be preferred over a copy, but often the law is silent on maintaining originals and so if copies are used in the normal course of business, this can be considered acceptable. In the common law system there has been a lot of confusion and inconsistency in court decisions over the status of a printed copy of an electronic original.

Two recent cases in the American courts have raised important issues relevant to this matter:

Armstrong v. Executive Office of the President. An injunction was sought in 1989 prohibiting the destruction of backup tapes of the email system used by the Reagan administration. Electronic versions of electronic mail were found to qualify as records and as such fell under the rules of the Federal Records Act. The plaintiffs asserted and courts concurred that an electronic mail message possesses unique characteristics that are lost when it is printed to paper. The case decided in favour of plaintiffs and upheld on appeal.

Public Citizen v. John Carlin. This case was brought against the National Archivist of the USA, challenging his authority to authorize the blanket destruction of electronic records. According to plaintiffs keeping paper copies of electronic records was not an adequate provision. According to NARA none of the electronic systems in place were able to meet NARA’s record-keeping requirements and they had no choice but to recommend that these records were printed to paper and captured in the traditional paper record-keeping system. The case was originally decided in favour of the plaintiffs in 1997, but reversed on appeal in 1999.

As has been detailed above, careful examination of the legal environment and archival implications is a first step. If there are no legal constraints, and if printing to paper is appropriate for managing certain documents it can be considered a supportive preservation method. Cases where there are no financial and organisational means available to provide for an alternative solution, or when no digital preservation solution is yet in place may warrant this technique.

Printing to paper is not a viable solution in any combination of the following cases:
() When there are any legal or archival constraints involved, as detailed above.
() When the prerequisites stated above are not fulfilled, i.e. when the documents are too complex or when the digital functionalities must be preserved.
() When there are too many files. In this case, financial considerations would strongly suggest implementing a digital archiving solution.
() When the digital documents are not born digital, but have been digitised. In this case, either the original paper documents or the digitised ones should be archived.

The dangers and shortcomings of printing to paper have been pointed out. With the necessary precautions taken into account, however, it might be considered as an alternative or partial solution for certain contexts.


United Nations Model Law on Electronic Commerce

MacNeil, Heather, Trusting Records: Legal, Historical, and Diplomatic Perspectives, Kluwer (2000)

A strong case for computer output on microfilm (but neglecting the legal implications) is made in: Young, Robert M., Availability & Preservation. Longterm Availability & Preservation of Digital Information. AIIM Industry White Paper on Records, Document and Enterprise Content Management for the Public Sector, 2002.

Submitted by kulc on 28 March 2003 at 12:21

Hi! We are about to study the destruction of electronic records after they're through their retention period. So, are you aware of any research works or guidelines on this topic, to make it a bit easier?

Answered by britished on 11 April 2003 at 10:04

This question relates to the actual destruction of electronic records once they have reached their scheduled date of destruction.

An electronic record, scheduled for destruction, must be disposed of in such a way that it cannot be retrieved. Disposal is the ability to identify, gain authorization, and completely purge a record from a computer system. Procedures for disposal of records must exist and be implemented consistently.

The record must be purged, as well as all metadata, index points, audit trails, and links to the record. Both primary and backup media must be included to ensure that the information contained on that media can not be reconstructed. Many public institutions must have accountable disposal procedures and keep track of what records were destroyed and how.

Possible methods of destruction (dependent on the information and the medium) include:
· Reformatting
· Rewriting
· Degaussing
· Electronic file shredding software
· Physical destruction of the storage medium

Most governments have legislation which governs the disposal and destruction of government records. Businesses should establish policies governing the destruction of records as part of their management of valuable information resources.

Further Information
BC Archives Policy 2-02: Destruction of Government Records
State Records Office of Western Australia, Disposal of State Records
State Records Authority of New South Wales, Destruction of Records- A Practical Guide (Presents a good overview of both principles and methods of destruction.)
National Archives of Australia, Protective Security Manual

Submitted by swissed on 14 January 2003 at 9:36

We are preparing to start a research project in physics. Since this will be state-funded and its results are expected to be of high value to other researchers as well, the reusability of our data is a crucial issue. Are there any guidelines or best practices we could use to achieve this goal? (Since the data will be transferred to a central repository, media deterioration should not be an issue.)

Answered by swissed on 16 January 2003 at 10:18

Scope of Question

This question relates to the basic steps needed to ensure data exchange and reuse, focusing on issues of logical rather than physical structure. (Metadata necessary for the management of the central repository are assumed to be created at transferral time and are not addressed here.)


The value of scientific data has long since been highlighted. They are often very difficult or impossible to reproduce (unique observations, high research costs, and other reasons). Thus, data reusability is a crucial feature in research projects. Apart from technical problems associated with the support media, this involves chiefly issues of logical structure. Fortunately, it is possible to greatly enhance data reusability by introducing a few basic measures.

A simple and concise guideline is the one established by R.B. Cook et al. (see references). This highlights seven best practices, namely: (1) choose descriptive file names; (2) adhere to consistent and stable file formats; (3) clearly define the parameters used; (4) organise data consistently; (5) perform basic quality assurance; (6) give the data sets comprehensive titles; and (7) provide clear and ample documentation in a non-proprietary format. These recommendations are straightforward and mostly easy to implement.
For recommendations that are more precise and detailed, but also more difficult to implement, one has to look at other studies that present comprehensive sets of best practices. The ESDIS Data Center Best Practices and Benchmark Report (2001) lists 35 best practices and discusses their concrete impact. And the two studies commissioned by the National Research Council – to cite only these – deal with several issues reaching from data creation to digital preservation properly.

After all, care for data exchange and reuse is the best premise for successful long-term preservation of scientific data. In the digital preservation community it is an axiom that long-term preservation must begin at data creation time.


COOK, Robert B; OLSON; Richard J.; KANCIRUK, Paul; HOOK; Leslie A.: Best Practices for Preparing Ecological and Ground-Based Data Sets to Share and Archive. 2000.
HUNOLT, G.; BOOTH, A.: ESDIS Data Center Best Practices And Benchmark Report. NASA, 2001.
Committee for a Pilot Study on Database Interfaces, National Research Council: Finding the Forest in the Trees: The Challenge of Combining Diverse Environmental Data. Selected Case Studies. Washington D.C., National Academy Press, 1995. ISBN 0 309 05082 0.
Steering Committee for the Study on the Long-term Retention of Selected Scientific and Technical Records of the Federal Government, National Research Council: Preserving Scientific Data On Our Physical Universe. A New Strategy for Archiving the Nation’s Scientific Information Resources. Washington D.C., National Academy Press, 1995. ISBN 0 309 05186 X.

Submitted by lspurdle on 13 January 2003 at 14:11

We are in the process of photographing several thousand museum objects at an archival standard.
At the moment images are kept both on the server and on CD-Rs.
Is it safe to keep these images on CD-R only, assuming that we had more than one copy and they were kept on differnt sites?
How long can CD-Rs be safely expected to last?

Answered by dutched on 11 March 2003 at 15:33

_Scope of Question_
Your question refers to the expected lifetime of optical media, specifically, write-once CD-ROMs. While the software format used for the data to be stored on the CDs needs to be addressed as well, this question purely focuses on hardware technology.

The lifetime of optical media varies greatly. The predicted physical lifetime varies between 50 and 100 years according to the manufacturers themselves. Tests by objective third parties revealed far shorter lifetimes, however. For all these tests it has to be kept in mind that they are based on methods for ‘accelerated ageing’ taking the short existence of optical media into account. Thus, these tests can only give theoretical results and estimations, no definite answers.
A recommendation that is often given is to migrate the data on CDs at least every 10 years to refresh media. However, these kinds of recommendations also vary. According to NARA (National Archives and Records Administration, USA), for example, the life expectancy of CDs can only be put to 3 to 5 years.

The lifetime of CDs largely depends on the quality of the CD and the storage conditions.
The quality of the CD is determined by the material it is composed of. This refers to components like the reflecting layer, the coating, and others. While the material used for rewritable CDs is generally not as durable as for non-rewritable media, no comprehensive experiments have been made up to now by an objective third party that analyse exhaustively the various components of the diverse types of CDs available.
Storage conditions have a great influence on the lifetime of a CD. Avoid storage areas in which temperature or humidity is high or fluctuates and handle the CDs with care.

However, the physical life-time of the CDs is rarely the constraining factor for ensuring the accessibility of the digital objects contained in them. Devices are needed to read the data available on the storage media. These peripherals and their connection to a system able to process the data need to be maintained along with the carrier. To date CDs are fairly wide spread, but DVDs could take over in no time and then CDs are essentially worthless if no backwards compatible DVD devices are available any longer.

In the absence of any ‘permanent’ solutions for archival storage of digital data, the use of CDs might be viable, provided - as described above - (a) the quality of the CDs is ensured, (b) the storage facilities are appropriate, (c) the maintenance of the devices is arranged, (d) a technology watch is put into place, and (e) periodic transfer of the digital objects to fresh storage media is organised. Furthermore and above this, measures have to be taken that the software formats of the digital objects stored on the disks are accessible in the future.

* Ross Harvey: The longevity of electronic media: from electronic artefact to electronic object. In Multimedia Preservation: Capturing The Rainbow, Brisbane, 27.-30. November 1995. National Preservation Office (NPO) Conference.

* John Van Bogart (National Media Laboratory): Storage Media Life Expectancies. Presentation at the Digital Archives Directions Workshop; June 22, 1998.

* The Proceedings of the AES 20th International Conference 2001 October 5-7 Budapest, Hungary - Archiving, Restoration, and New Methods of Recording", Co- chairs Éva Arató-Borsi and Dietrich Schüller, Audio Engineering Society 2001.
-> Drago Kunej: Instability and Vulnerability of CD-R Carriers To Sunlight (Results of a Simple Experiment in Everyday Work Environment)
-> Practical Experience with Long- Term Archiving of Data on CD-R

* Hartmut Weber: Opto-electronic storage - an alternative to filming? The Commission on Preservation and Access - Newsletter 53, February 1993.

* Media Sciences;

Submitted by David Kenny on 10 January 2003 at 13:55

We are a NOF digi project (FenPast) and use a Betterlight Super 6k scanning back with a Cambo Ultima 5x4 copy camera. We need an A3 scanner to help us increase our productivity without compromising scan quality. We are are thinking of buying a Fuji Finescan 2750 A3 scanner but are unsure as to whether it is worth £3000+ more than a Umax PowerLook for our needs. Any advice would be welcome.

Answered by swissed on 21 January 2003 at 14:34

Scope of Question

This question relates to the wider issue of choosing equipment for scanning.


The choice of the most suitable technical equipment for a digitisation project is largely determined by this project’s objectives and requirements. The choice of scanner should be based on the materials to be scanned and their intended purpose.

Here are some considerations that should influence the decision:

· Other hardware equipment: scanner properties should match the properties of other equipment used. It is not sensible that the scanner resolution excels the memory and processor capabilities of the computer used. Similarly, limited storage sizes suggest employing a smaller scanning resolution.

· Type of media to be scanned: some media allow for automated or semi-automated scanning. Others have limited possibilities, dictated by their materials: books may suffer when put on flatbed scanners, as do glass negatives on glass surfaces. The warming caused by the light source should be considered. Fragile materials need careful handling, which also slows down the scanning speed.

· Quality of media to be scanned: high resolution in scanning is only sensible for high-resolution originals. It is not reasonable that the scanner resolution exceeds the originals’ resolution.

· Scanning speed: automatic or semi-automatic scanners can handle large quantities of originals, if the materials are suitable. The highest throughput can be achieved by drum scanners and hybrid drum scanners.

· Quality and use of products: the quality needed for the digitised images is determined by the use that will be made of them. Archival copies of photographic images should be made at 500 dpi at least, using 8-bit greyscale or 24-bit colour. Slides and negatives call for a much higher resolution. However, if fast access to the images is paramount, a smaller resolution is desirable, which also allows for a faster throughput.

· Metadata: sufficient metadata collection and creation are crucial for the use and storage of any digitised images. Check if the scanner allows for automated metadata capturing.

· Long-term preservation considerations: archival formats like TIFF should be preferred whenever digitised images are to be kept for any length of time. These are standard formats and allow for conversion to other formats.

Besides these considerations, make sure that the scanner’s analogue (optical) and digital (CCD, AD-conversion) imaging qualities match. Of course, the scanner needs to be calibrated, and its dark current, noise, and sensitivity should be recorded regularly. These records belong to the scanning metadata and must be stored along with the images.

After considering and weighting these aspects the properties of the scanner needed should become clearer. It is only then that the question whether these are worth their price or whether small concessions are acceptable can be answered.


* ERPANET Toledo Workshop Report: Digitisation, Conservation, and Quality Control. May 2002

* NARA Guidelines for Digitizing Archival Materials for Electronic Access, January 1998.

* The Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH):
The NINCH Guide to Good Practice in the Digital Representation and Management of Cultural Heritage Materials. October 2002.

Previous | Next