Use It or Lose It! Preserving Your Digital Documents

Rob Spindler, University Archivist
Arizona State University Libraries ECURE 2004

1

Eight Challenges of Electronic Document Preservation

• A review of fundamental issues
• Applies to desktop and enterprise level
• All eight items must be addressed for success

2

Eight Challenges — Storage Media

1. Physical Degradation of Storage Media

• CD's last 30-100 years, BUT
  • Temperature/Humidity
  • Pollutants
  • Improper Handling
    • Scratches, bent
    • Label adhesives/Inks

3

Eight Challenges — Storage Media

Physical Obsolescence of Storage Media

• 5.25" floppy discs
• Obsolete backup tape cartridges
• Obsolete optical discs/cartridges

4

Eight Challenges — Storage Media

Obsolescence of Media Recording or Playback Devices

• Data storage efficiency increases 10X-50X every 10 years
• Recording device production life about 10 years.
• Mfg. product support 5-7 yrs. after production termination

5

Eight Challenges — Storage Media

Four Types of Recordable DVD

• DVD-R
• DVD-RW (Pioneer device)
• DVD-RAM — Type 1 & 2 (IBM, Gateway)
• DVD+R / DVD+RW (Hewlett-Packard, Dell)

6

Eight Challenges — Software

Migration Failure

Three kinds of migration:

• Migration to new release of same software.
• Migration to new software
• Migration to software independent format (e.g. ASCII Text)

7

Eight Challenges — Software

Migration Failure — (Software, Operating System or Encoding Incompatibility)

• New software release won't run file from old release.
• New software release opens old file but contents are corrupted.

8

Eight Challenges — Software

Migration Failure — (Software, Operating System or Encoding Incompatibility - Continued)

• Software is not compatible with operating system.
• Display with one internet browser doesn't look the same in a different browser..

9

Eight Challenges — Software

Migration Failure — (Software, Operating System or Encoding Incompatibility - Continued)

• Browser does not support encoding standard used (XML).
• Proprietary codes from HTML editing packages do not convert to new HTML applications.

10

Eight Challenges — Human Factors

Human Error / Vandalism

• Accidental or malicious deletion
• Website is vandalized
• Virus protection failure

11

Eight Challenges — Backups

Backups and Snapshots

• Backups lost or overwritten
• Wrong files backed up
• Incomplete snapshot: Video server backed up, web server not

12

Eight Challenges — Metadata

Lost or Absent Metadata

Data about Data

Two Types:

• Administrative
• Descriptive

13

Eight Challenges — Metadata

Administrative Metadata

• Hardware and Software Requirements
• Maintenance & Migration History
• Version Control
• Data Dictionaries

14

Eight Challenges — Metadata

Descriptive Metadata

• Media Labels
• File Lists
• Electronic Descriptions:
  • File folder titles
  • Filenames
  • File Header — Dublin Core

15

Eight Challenges — Metadata

Linkage/Addressing Failure

• URL in document changes
• Linkage between description and digital object fails

16

What to Do? Recommendations...

1. Actively manage your files!
2. Select Carefully!
3. Refresh and/or migrate to new storage media on a regular schedule
4. Consider redundant and/or offline storage

17

What to Do? Recommendations...

5. If you can, use standards:
   • ASCII text
   • Unicode (supports foreign languages)
   • For HTML web documents, use W3C standards and verification service
   • For encoded text: TEI, XML

18

What to Do? Recommendations...

6. If you must, select robust software:
   • Use widely adopted software with "large installed base"
   • Look for proven export capabilities
   • Ask about backwards compatibility

19

What to Do? Recommendations...

7. Plan for migration and emulation:
   • Migrate legacy and active content:
     • to new release of same software
     • to compatible different software
   • Test migrations w/sample data
   • Retain obsolete software and files

20

What to Do? Recommendations...

8. Address Human Factors:
   • Provide adequate training
   • Conduct audits and quality control checks
   • Create, maintain and document procedures
   • Maintain secure systems and virus protection

21

What to Do? Recommendations...

9. Create and Maintain Metadata:
   • Label your storage media
   • Create descriptive filenames and file folder titles
   • Make Data Dictionaries — consider paper or offline
   • Use Dublin Core for Webstuff

22