IS&T Archiving Conference CURATEcamp 2013/formats

Problem of Unidentifiable formats (e.g. research data from sensors)

• They don’t consult them before buying sensors
• Some things you can do about it
  • Expose problem files to tool developers
  • More of a knowledge base problem than tool problem - contribute to format registries
• Hard problem to identify binary and proprietary formats - not many people working on it
• Undocumented formats - not much hope

Normalization? - good or bad

• Pro: Reduce number of formats
• Con: Lose authenticity?
• Content situational (email)
  • Need Tools to identify it and normalize it (some scientific data not possible)
  • When to consider it
    • Content warrants it (expensive to produce)
    • Format warrants it
      • Email - Difficult to preserve native formats long-term
• Loss during transformation
  • We might be losing information already
    • RAW to other formats (DNG - an attempt to normalize RAW formats)
    • DNG can embed the original format in it
    • inherent problem with RAW - it needs to be interpreted
    • Nice to have when you want to go back to the RAW file
• Transformations in general - Losslessness?