Digital Preservation: The Migration Method - Research Paper Example

Digital Preservation: The Migration Method In the past, much of archiving physical objects involve their maintenance in their original forms but, unfortunately, this is no longer necessary or sufficient for objects that have been digitized. One of the greatest challenges in digital archiving is preservation. The difficulties and problems that librarians and archivists must confront in regard to preserving digital information are demonstrated in the Australian government’s Preserving Access to Digital Information (PADI) site. To quote: Compared with an object in a museum which may lay undisturbed for years in a storeroom or a book on a shelf, or even Egyptian hieroglyphics carved on the wall of a tomb, digital information requires much more active maintenance. If we want access to digital information in the future, we must plan and act now. There is too much information being churned out of traditional media and from the web. In particular, information from governmental, cultural and in the university settings are continually being generated in digital form and that if these materials are lost, then much documentary heritage will be absent from the historical record. In August 2006, an important example of digital disaster was highlighted and naturally served as a wakeup call for stakeholders. Tapes with very high-quality pictures of the first moon walk in 1969 have been reported missing and that to add insult to injury, the only remaining tape drive that can read these tapes is scheduled to be retired within the year. The missing tapes have a much higher quality picture than what was seen on, and recorded for, television and so the loss is incalculable. (Raisinghani, p. 406) Two principal approaches to preserving digital objects have been pursued particularly intensively during the last years: migration and emulation. While emulation is the strategy operating on the environment of the objects to make sure that is everything that is needed for rendering those object is in place, migration transforms them into new representations that are considered to be safer and better-suited for long-term preservation. In the following, we will explore the process of migration; discuss its various forms – their advantages and disadvantages, and the problems and challenges that this digital preservation strategy must contend with. BACKGROUND One of the characteristics of computing technology is that it is highly dynamic and rapidly evolves every day. New versions of hardwares and softwares are continually introduced to the market, significantly affecting the way digital data are stored and preserved. When a new operating environment, for instance, is introduced, not only the physical storage medium is changed but that an entire digital environment must be transferred. This process is called migration or the periodic transfer of digital materials from one hardware/software configuration to another or from one generation of computer technology to the next is the current favorite strategy for preserving electronic data. According to Henry Gladney (2007), the periodic migration of digital objects is characterized by two distinct principles. The first, faithful copying of bit-strings from one substratum to a successor substratum is simple and reliable; secondly, copying with change of format from a potentially obsolete representation to a more modern replacement, is a complex task requiring highly technical expertise. (p. 9) As the process focuses on the digital object itself in its processes, digital migration is seen as the most viable method of long-term digital preservation. The reason is that its purpose is to “preserve the integrity of digital objects and to retain the ability for clients to retrieve, display, and otherwise use them in the face of constantly changing technology.” (Gladney p. 8) In 2004, an international survey from among national libraries, state libraries and consortia, archives, museums, and other organizations from 13 countries reported that 92 percent preferred migration over other digital preservation strategies. (Harvey 2005, p. 102) FORMS There are three forms of migration: migration at obsolescence, format normalisation and migration on demand. Migration at Obsolescence/Format Migration This process is the migration of digital objects from one version of a format to a later version of the same format. In this case, not only the data carriers are exchanged, but also the format of digital data. The process of changing an uncompressed TIFF to uncompressed JPEG2000 demonstrates this, or when a Word 2000 documented is converted to Word 2007 format or a JPEG image to a JPEG2000 format.. This migration type is difficult for several reasons: 1. There is a need to guarantee that no data loss occurs with the format conversion. 2. The proof that the copy process succeeded becomes more difficult. The file in the new format has to be converted back to the old format and then compared with the original file. The comparison must be made in logical level (for example, comparing the pixel values in the case of images) and not on a bitwise basis of the resulting files. The rationale behind this tedious process is that the structure of the files may differ, even if they represent exactly the same content. Two TIFF files, for example, may differ on a bit level but present identical images with identical information content. 3. Using more recent versions of softwares, even with the original formats, may present the preserved documents with characteristics they did not, and perhaps could not, have had. (Thibodeau) 4. Product lines and, hence, the migration path, could be terminated. A logical strategy in regard to format migration is for a digital object to be created or migrated to the format with the most information, in the most open format to assure the least risk of obsolescence. The good thing about format migration, however is that format conversions are only necessary only if the format becomes obsolete unlike, say a bit-stream migration, wherein conversion is necessary every 5 years. Format Normalisation This is the process of migration to a standard format or restricted range of formats and carried out at time of ingest. The object of migration here is a better long-term preservation prospects – usually non-proprietary open formats. For example, a textual document such as a WordPerfect document, could be reduced to plain ASCII. According to Thibodeau, this conversion is eminently practicable and appropriate in cases where essential characteristics to be preserved are the textual content and the grammatical structure. How format normalisation minimizes heterogeneity and its focus on future migration steps simplify long-term administration. For example, a proper detachment of the digital object from its original media decreases efforts for monitoring interpretation aids that are necessary such as for media-specific formats. This is further demonstrated in how normalisation can happen at two levels: File formats: An archive might prefer to store all raster images in TIFF, for instance, and convert JPEG or GIF images into that format. Controlling the number of file formats will reduce the complexity of format monitoring and migration. Document formats: Many e-journal publishers encode article content in SGML or XML (or plan to do soon). Most publishers create their own DTD (or modify an existing DTD) to suit their specific needs and delivery platforms. An archive might choose to normalize all such marked-up documents into a common DTD, reducing the complexity of documentation, migration and interface software. (Lakshmi and Jindal, p. 80) The inadvisability of using standard format was demonstrated in a case cited by the Cedars Project: An organization running an IBM mainframe in the 1980s wanted their data to be accessible beyond the life of its current system. The textual parts of this data are held in EBCDIC. But the replacement systems installed in the 1990s operate using ASCII character set. The organization is faced with a problem: to convert the document archive to ASCII, or to retain it in EBCDIC. The organization risked permanent information loss in the arbitrary conversions where the EBCDIC character has no equivalent in ASCII. As with other migration strategies, the format normalization strategy also involves the risk of information loss. A normalization format example could be found in the National Archives of Australia website (http://www.naa.gov.au/records-management/secure-and-store/e-preservation/at-NAA/index.aspx ). Here, images are normalised to png, the audio files to bwav format, while other documents such as those made in MS Office and OpenOffice.org are converted to XML. Migration on Demand Migration on demand simply means that migration from an older to the most current version occurs upon access and only for access purposes. This migration process, however, requires an emulator that will migrate the data object to a format that is usable in the current hardware/software/OS environment. This raises a large number of issues. The most significant of this is that because of the rapid rate at which storage media are being developed, the emulation of a given storage environment would be a formidable challenge. Presently, there is no evidence of projects wherein this category has been implemented but that a proof of concept is in the CAMiLEON (Creative Archiving at Michigan and Leeds: Emulating the Old and New) project being developed in the University of Michigan. According to the project website, “a small tool has been developed to investigate and demonstrate Migration on Request as an approach to digital preservation, and to identify any challenges associated with implementing this form of migration.” This is called the Vector Graphic Migration on Request tool wherein vector graphic files define images in terms of their components. The site how the tool works: For example, a stick man could be formed from a circle and a few straight lines. Each part is generally termed an ‘object’ or ‘element’. A vector graphic file would store such a picture in terms of each element’s coordinates and attributes (such as fill colour, line colour, line thickness etc.) According to the CURL Exemplars in Digital Archives or the Cedars Project not only does this Migration on Request/Migration on Demand avoid repeated migration of the entire archive as technology progresses, but may well cut out steps that relate to formats that have come and gone in the interim between preservation and access. (UK Web Archive) Since this method is still under development, it is still difficult to determine its actual advantages and disadvantages. MIGRATION CHALLENGES Cost Migration, as a strategy in the preservation of digital objects has been endorsed by many libraries and archives, as the most practical approach, at least for the short and medium term. However, the costs of migration may, in the long run, exceed those costs that would be incurred by technology preservation. Along with the digital preservation strategy – emulation – migration requires a large commitment in resources up-front and over a long-term: ongoing migration requires intensive cyclical work to convert objects in obsolete formats to current formats and that the work increases as the digital collection grows. (Heslop, Davis and Wilson 2002, p. 12) Because preservation of digital objects has an ever-increasing cost, digital libraries and archives are in dire need for the right policies and strategies to store and preserve information especially in the long-term. It remains impossible to estimate the costs of digital preservation over the whole life of digital objects. The expectation is that the amount of data to be stored will increase quite spectacularly on a number of fronts: the size of a digital object will increase as it acquire and incorporates new features and functionality, the number of objects will continue to grow, and migration programs will also increase its data storage requirements each time a migration process is completed. It is important to underscore that a life-cycle of information is not only important in regard to preservation and access, but that it is also pivotal in the full business model of digital libraries and archives. Once the digital archiving systems are set up, the implementation of digital preservation strategies becomes relevant. Since migration is expected to be an expensive strategy, the level of funding will also become important. However, a number of national libraries are starting to engage themselves in a series of collaborative efforts in regard to developing registries of file format and representation information for preservation planning. (Verheul, p. 54) This is, at least, an area where there is likely to be significant cost reductions and preservations benefits. The collaborative activity, however, requires complicated governance and business model issues and that they may need to be resolved before viable services could be introduced. Nonetheless, uncertainties about costs may hinder progress of digital migration efforts. Risk Some attributes of the digital object may be damaged and subsequently, lost, during the conversion process, therefore the performance or fidelity to the original source may not be equivalent after migration. The level of data loss through migration, wrote Heslop, Davis and Wilson, depends on the number of preservation treatments applied to the record, the choice of process, the new data format, the level of human intervention and post-migration descriptive work. (p. 12) Furthermore, the strategy does not always allow the making of an exact digital copy or replica of the digital objects (as hardware and software change) while maintaining the compatibility of the digital object with the new generation of technology. Preservation Management The migration model requires following several rules of constant care so as the data will not be lost. Michael Peres outlined these rules as follows: There is the rule on redundancy wherein data must be kept in several copies in different locations. For all the data files, checksums should also be calculated and archived with the data files, which allows for checking data files at any time for aging-related changes or errors. Every 12 to 24 months, the data must be proofread and the checksums compared as well. When errors are detected, the data must be migrated immediately. Migrations have to be planned in advance, including financing. A format migration is advised if a new file format becomes standard and the conversion can be done without loss of data. Every step has to be documented in detail and all media must be labeled properly. (p. 364) The National Archives of Australia has also developed a process called Designing and Implementing Recordkeeping System (DIRKS) – an eight-step process defining the best practice standards and guidelines. The DIRKS methodology is a basic waterfall requirements process that is designed to “help each organization to determine such requirements and put in place procedures to reassess these needs over time.” (NAA 2003) Authenticity Another problematic area in regard to migration is determining authenticity. Presently, the current authentication technique for authenticity is the digital signature. Here, a digital object or document is authenticated by encrypting it with a private key of the authenticating author. This key would function as a tool to confirm the authenticity of the object. Unfortunately, it cannot be passed on when the object are being migrated and converted and that re-signing would involve access to a private key that may well not be available anymore. (Holsworth 2007, p. 18) Even if new technology would be able to re-sign the digital code, it would further complicate the migration process as the encryption would form part of the routine process of copying from old media to current media. CONCLUSION Unarguably, the digital migration approach is not without flaws. It is labour intensive, time-consuming, expensive, error-prone, risky (as its processes are still marred by lost and corrupted information, and, non-scalable, for the process can require new solutions for each new format. But migration has so far been the only serious candidate for digital preservation of large scale archives. According to Thibodeau, the ideal preservation system would be a neutral communications channel for transmitting information to the future and this channel should not corrupt or change the messages transmitted in any way. Migration – in all its forms – appears to be the more productive option in line with this criteria today and the future, since at least it moves the problem of managing bits rather than atoms. The method allows us to separate components and interpose levels of redirection to take account of changes in components and attributes. Indeed, one could say that the search for the perfect solution in digital archiving is premature, unrealistic and could, in fact, be counter-productive. In addition, problems in digital preservation is not exclusively technical as they may involve difficult social, political, and organizational questions of authenticity, ownership and responsibility as well. Furthermore, there is still no single universal strategy that could provide all-encompassing solution to the digital preservation needs. In regard to the kinds of migration strategies, they are not mutually exclusive because there are different kinds of digital materials as well as their access requirements. According to Ross Harvey none of the strategies are complete, however, they allow us to identify some of the principles that are important and to make some judgments about the applicability and viability, particularly in the longer term. (p. 114) Digital preservation is expected to evolve and that it must. “If the reservation solution,” wrote Thibodeau, “cannot grow and adapt to continuing changes in the nature of the problem and continuing escalation of user demands, the ‘solution’ will in short order become part of the problem; that is, it will itself become obsolete.” In its development, migration and its various methods should finally be able to offer a high level of safety and quantity, volume of storage, ease of access and other attributes at relatively low cost. It should allow these opportunities to be changed for a given document, type of document or any other recorded object at any time in the future. Ideally, migration should also be able to provide single-step access to all information without the requirement for multiple layers of encapsulation to be detached to access older documents and at the same time, allowing the contents of a digital document to be extracted for conversion into the current vernacular, without losing the original form of the document. References CAMiLEON. (not dated). The Vector Graphic Migration on Request Tool. CAMiLEON. Retrieved June 26, 2009, from < http://www.si.umich.edu/CAMILEON/reports/mor/vector.html> Gladney, H. (2007). Preserving Digital Information. Springer. Harvey, R. (2005). Preserving Digital Materials.K.G. Saur Munchen. Heslop, H, Davis, S. and Wilson, A. (2002).An Approach to Digital Records. National Archives of Australia. Holworth, D. (2007). Preservation Strategies for Digital Libraries. DCC Digital Curation Manual, S. Ross, M. Day (eds) Retrieved June 26, 2009, from http://www.dcc.ac.uk/resource/curation-manual/chapters/preservation-strategies-digital-libraries Lakshmi, V. and Jindal, S.C. (2007). Digital Libraries. Gyan Publishing House. Lyall, J. (2009). PADI: Preserving Access to Australian Information and Cultural Heritage in Digital Form. National Library of Australia. http://www.nla.gov.au/nla/staffpaper/lyall3.html Peres, M. (2007). Focal Encyclopedia of Photography. Focal Press. Thibodeau, K. (not dated). Overview Of Technological Approaches to Digital Preservation and Challenges in Coming Years. Clir.org.Retrieved June 22, 2009, from UK Web Archive. Cedars Guide to Digital Preservation Strategies. UK Web Archive. Retrieved June 22, 2009, from http://www.webarchive.org.uk/wayback/archive/20050409230000/http://www.leeds.ac.uk/cedars/guideto/dpstrategies.html Verheul, I. (2006). Networking for digital preservation: current practice in 15 national libraries. Walter de Gruyter. Read More