diff --git a/iep-0002/index.html b/iep-0002/index.html index 44e3140..055325d 100755 --- a/iep-0002/index.html +++ b/iep-0002/index.html @@ -271,6 +271,13 @@ +
  • + + + 9. Bibliography + + +
    • @@ -478,6 +485,13 @@ +
  • + + + 9. Bibliography + + +
    • @@ -524,7 +538,7 @@

    I Updated: -2023-12-28 +2024-01-04 @@ -706,7 +720,7 @@

    8. Metadata extraction -

    Bibliography

    +

    9. Bibliography#

      @@ -726,7 +740,7 @@

      8. Metadata extraction -2023-12-28 +2024-01-04 diff --git a/iep-0008/index.html b/iep-0008/index.html index a143131..3c35f1e 100755 --- a/iep-0008/index.html +++ b/iep-0008/index.html @@ -256,6 +256,13 @@ +
    1. + + + 7. Bibliography + + +
      2. @@ -364,6 +371,13 @@ +
    2. + + + 7. Bibliography + + +

    @@ -410,7 +424,7 @@

    I Updated: -2024-01-03 +2024-01-04 @@ -479,12 +493,12 @@

    6. Conformance -

    Bibliography

    +

    7. Bibliography#

    1. -

      Collet, Yann. xxHash: Extremely fast hash algorithm. +

      Collet, Yann. xxHash: Extremely fast hash algorithm.
      Accessed July 2022, available at https://cyan4973.github.io/xxHash/ ↩

    @@ -494,7 +508,7 @@

    6. Conformance -2024-01-03 +2024-01-04 diff --git a/iep-0009/index.html b/iep-0009/index.html index 37a6ce3..a0280bd 100755 --- a/iep-0009/index.html +++ b/iep-0009/index.html @@ -204,11 +204,74 @@
  • + IEP-0009 - Instance-Code +
  • @@ -261,6 +324,61 @@
    @@ -290,7 +408,7 @@

    2023-12-28 +2024-01-04 @@ -317,12 +435,97 @@

    ISO/DIS 24138

    • +

    1. General#

    +
      +
    1. The Instance-Code shall be a checksum for any kind of data regardless of its media type.
      2. +
    2. The Instance-Code shall match identical files and indicate data transmission errors or data manipulation.
      3. +
    3. The Instance-Code shall be the prefix of a merkle root of a merkle tree as illustrated in Figure 11.
      4. +
    4. The cryptographic hash algorithm used for Instance-Code calculation shall be updated if a security weakness is discovered.
      5. +
    5. An update shall be indicated by a higher version number of the Instance-Code header.
      6. +
    +
    +

    Figure 11 - Example of a merkle tree +

    +
    Figure 11 - Example of a merkle tree
    +
    +
    +

    NOTE 1

    +

    Depending on security and uniqueness requirements, applications can generate and store +Instance-Codes up to a length of 256 bits in steps of 32 bits. Longer variants of an +Instance-Code are extensions of the shorter versions (which function only as checksums) and +remain compatible while improving the security of the hash.

    +
    +
    +

    NOTE 2

    +

    The use of tree-based hashing supports efficient containment-proofs, verified streaming and +verification of data from trusted and untrusted sources. A data delivery application can +provide cryptographic proofs for partial data and streams such that the receiving application +can incrementally verify chunks of data against a given Instance-Code.

    +
    +

    2. Format#

    +

    The Instance-Code shall have the data format illustrated in Figure 12.

    +
    +

    Figure 12 Data format of the Instance-Code +

    +
    Figure 12 - Data format of the Instance-Code
    +
    +
    +

    EXAMPLE 1: 64-bit Instance-Code in its canonical form:

    +

    ISCC:IAAZ3NGA3HTIYUQD

    +
    +
    +

    EXAMPLE 2: 256-bit Instance-Code in its canonical form:

    +

    ISCC:IADZ3NGA3HTIYUQD3SGC737FF6S5KRTRXY5DEU7ANCEMVTT4MDS2OQY

    +
    +

    3. Inputs#

    +

    The input for calculating the Instance-Code shall be the bytes of a file, without reference to +their meaning or structure.

    +

    4. Outputs#

    +

    Instance-Code processing shall generate the following output elements:

    +
      +
    1. iscc: the Instance-Code in its canonical form (required);
      2. +
    2. datahash: a cryptographic hash of the input file encoded as multihash (optional);
      3. +
    3. filesize: the size of the input file in bytes (optional).
      4. +
    +

    5. Processing#

    +

    An ISCC processor shall calculate the Instance-Code as follows:

    +
      +
    1. Apply the tree-based cryptographic hash function to the file input using the appropriate number of bytes for the output digest.
      2. +
    2. Use the hash digest of output as the ISCC-BODY of the Instance-Code.
      3. +
    +

    For further technical details see source-code in modules +code_instance.py of the +reference implementation.

    +
    +

    NOTE

    +

    The merkle tree chunk size is defined by the BLAKE3 1 algorithm to be 1024 bytes.

    +
    +

    6. Conformance#

    +

    An implementation of the Instance-Code algorithm shall be regarded as conforming to the standard +if it creates the same Instance-Code as the reference implementation for the same data input.

    +
    +

    NOTE

    +

    The ISCC reference implementation uses the open source BLAKE3 1 library for merkle tree +hashing and appropriate use of this software will generate the same codes as the reference +implementation.

    +
    +

    7. Bibliography#

    +
    +
    +
      +
    1. +

      O’Connor, J., Aumasson, J.P., Neves, S., Wilcox-O’Hearn, Z., BLAKE3: one function, fast everywhere. +Version 20211102173700, accessed July 2022. +Available at https://github.com/BLAKE3-team/BLAKE3-specs/blob/master/blake3.pdf ↩↩

      +
      2. +
    +