1. Colorado Springs Producer-Archive Interface Specification Status of standardisation project Main characteristics, major changes, items pending

2. PAIS – 16-18 January 2007 2 Contents 1. Objective 2. Basic principles 3. Main changes 4. Items pending 5. Conclusion

3. PAIS – 16-18 January 2007 3 Objective ■Implementation standard for Producer – Archive interface  Conformity with the PAIMAS  Conformity with the XFDU ■Aimed mainly at Formal Definition and Transfer and Validation Phases

4. PAIS – 16-18 January 2007 4 Basic principles - The Plan of Objects to be Transferred  The term ‘POT’ may change following the current discussions ■The Plan of Objects to be Transferred (POT)  A Transfer Object is a structured and organized set of one or more Data Objects that are to be transferred to the Archive. (There may be multiple Transfer Objects of the same Transfer Object Type.)  Descriptors describe a Transfer Object Type or a Collection of Transfer Object Types.  Descriptors are related to each other by parent-child relationships, and specify relationships between the Data Objects corresponding to these types.  Descriptors are derived from one or more Models.  The set thus structured makes up the POT.

5. PAIS – 16-18 January 2007 5 Basic principles - SIP ■The Submission Information Packages (SIPs)  These are packets used for transferring data: a SIP must include one or more Transfer Objects.  All the components of a Transfer Object must be in the same SIP.  Packaging of SIPs is based on the XFDU packaging standard.  We will thus have a generic SIP Model derived from the XFDU scheme.

6. PAIS – 16-18 January 2007 6 Basic principles – SIP and XFDU ■A SIP consists most often of a single XFDU; however, if because of the volume of data to be transferred, the objects sharing a given Descriptor cannot be sent in a single XFDU, they will then be sent in a group of linked XFDUs.  Some attributes necessary for a SIP have been identified:  Producer_id  Project_id  Sip_type_id  And for each sub-packet corresponding to a Transfer Object:  Descriptor_id  Last_object  Updated_object  Lou Reich and his team are working on a technical solution to take these attributes into account.

7. PAIS – 16-18 January 2007 7 PAIS main dates ■November 2006: proposal for a draft Red Book (current version in progress) PAIS split into 2 documents:  The standard itself including the concepts definition and their implementation.  A tutorial explaining how to use the standard including examples. Main changes from the previous version? October 04October 05September 06February 06June 06 draft White Book versions 12345 ■Septembre 2004: concept paper

8. PAIS – 16-18 January 2007 8 Main changes: new organization of the content ■1. Introduction ■1.1Purpose and scope ■1.2Applicability ■1.3Rationale ■1.4 Conformance ■1.5Document structure ■1.6Definition ■1.6.1Acronyms and abbreviations ■1.6.2Glossary of terms ■1.7 References ■2 Overview ■2.1 Formally describing Data Objects for Transfer ■2.2 SIP creation and validation ■3 Descriptions of Producer Data ■3.1 Abstract Descriptor Models definition ■3.1.1 Transfer Object Descriptor ■3.1.2 Collection Descriptor ■3.2 Concrete Descriptor Models implementation ■3.2.1 Transfer Object Descriptor Model ■3.2.2 Collection Descriptor Model Comment: this is the table of contents for the body of the documentg (without classic introductive sections and annex) ■3.3 Abstract Specialization of the Descriptor Models ■3.4 Concrete Specialization of the Descriptor Models ■4 Descriptions of constraints on types of SIPs ■4.1 Abstract Constraints on types of SIPs ■4.1.2 SIP grouping constraints definition ■4.1.3 SIP sequencing constraints definition ■4.2 Concrete constraints on types of sips ■4.2.1 SIP grouping constraints implementation ■4.2.2 SIP sequencing constraints implementation ■5 SIP Model ■5.1 Abstract SIP Model ■5.1.1 Descriptors and SIPs ■5.1.2 Sip Model ■5.2 Concrete SIP Implementation ■5.2.1 XFDU implementation of SIP ■5.2.2 SIP Specialization ■ANNEX A ■A.1 Generic Transfer Object Descriptor Model ■A.2 Generic Collection Descriptor Model ■A.3 Generic SIP Model ■ANNEX B : Informative references

9. PAIS – 16-18 January 2007 9 Main changes: new organization of the content ■2 Overview ■2.1 Formally describing Data Objects for Transfer ■2.2 SIP creation and validation ■3 Descriptions of Producer Data ■3.1 Abstract Descriptor Models definition ■3.1.1 Transfer Object Descriptor ■3.1.2 Collection Descriptor ■3.2 Concrete Descriptor Models implementation ■3.2.1 Transfer Object Descriptor Model ■3.2.2 Collection Descriptor Model ■3.3 Abstract Specialization of the Descriptor Models ■3.4 Concrete Specialization of the Descriptor Models ■4 Descriptions of constraints on types of SIPs ■4.1 Abstract Constraints on types of SIPs ■4.1.2 SIP grouping constraints definition ■4.1.3 SIP sequencing constraints definition ■4.2 Concrete constraints on types of sips ■4.2.1 SIP grouping constraints implementation ■4.2.2 SIP sequencing constraints implementation ■5 SIP Model ■5.1 Abstract SIP Model ■5.1.1 Descriptors and SIPs ■5.1.2 Sip Model ■5.2 Concrete SIP Implementation ■5.2.1 XFDU implementation of SIP ■5.2.2 SIP Specialization

10. PAIS – 16-18 January 2007 10 Main changes: new organization of the content ■2 Overview ■2.1 Formally describing Data Objects for Transfer ■2.2 SIP creation and validation ■3 Descriptions of Producer Data ■3.1 Abstract Descriptor Models definition ■3.1.1 Transfer Object Descriptor ■3.1.2 Collection Descriptor ■3.2 Concrete Descriptor Models implementation ■3.2.1 Transfer Object Descriptor Model ■3.2.2 Collection Descriptor Model ■3.3 Abstract Specialization of the Descriptor Models ■3.4 Concrete Specialization of the Descriptor Models ■4 Descriptions of constraints on types of SIPs ■4.1 Abstract Constraints on types of SIPs ■4.1.2 SIP grouping constraints definition ■4.1.3 SIP sequencing constraints definition ■4.2 Concrete constraints on types of sips ■4.2.1 SIP grouping constraints implementation ■4.2.2 SIP sequencing constraints implementation ■5 SIP Model ■5.1 Abstract SIP Model ■5.1.1 Descriptors and SIPs ■5.1.2 Sip Model ■5.2 Concrete SIP Implementation ■5.2.1 XFDU implementation of SIP ■5.2.2 SIP Specialization New organization of the standard: Previous version: 2 separate parts, one for the definition of the concepts and one for the implementation. current version: Implementation immediatly follows the definition of each concept

11. PAIS – 16-18 January 2007 11 Main changes: new organization of the content ■2 Overview ■2.1 Formally describing Data Objects for Transfer ■2.2 SIP creation and validation ■3 Descriptions of Producer Data ■3.1 Abstract Descriptor Models definition ■3.1.1 Transfer Object Descriptor ■3.1.2 Collection Descriptor ■3.2 Concrete Descriptor Models implementation ■3.2.1 Transfer Object Descriptor Model ■3.2.2 Collection Descriptor Model ■3.3 Abstract Specialization of the Descriptor Models ■3.4 Concrete Specialization of the Descriptor Models ■4 Descriptions of constraints on types of SIPs ■4.1 Abstract Constraints on types of SIPs ■4.1.2 SIP grouping constraints definition ■4.1.3 SIP sequencing constraints definition ■4.2 Concrete constraints on types of sips ■4.2.1 SIP grouping constraints implementation ■4.2.2 SIP sequencing constraints implementation ■5 SIP Model ■5.1 Abstract SIP Model ■5.1.1 Descriptors and SIPs ■5.1.2 Sip Model ■5.2 Concrete SIP Implementation ■5.2.1 XFDU implementation of SIP ■5.2.2 SIP Specialization  Synthetic diagram identifying the objects  Synthetic diagram showing the whole process New organization of the standard: Previous version: 2 separate parts, one for the definition of the concepts and one for the implementation. current version: Implementation immedialtly follows the definition of each concept

12. PAIS – 16-18 January 2007 12 Main changes: overview Collection Descriptor ABC Collection Descriptor C Collection Descriptor AB Transfer Object Descriptor (TOD) A Described by Digital Objects of type A 01000010001 Transfer Object Descriptor (TOD) C Described by Transfer Object Descriptor (TOD) B Described by SIP Content Type 1 TOD A TOD B SIP Content Type 2 TOD C Child-Parent Relationship Relationship as Labeled SIP Sequencing Constraints group one - SIP Content Type 2 - SIP Content Type 1 LEGEND Referenced in Digital Objects of type B 01000010001 Digital Objects of type C 01000010001 Referenced in Identified in Digital Objects to be Transferred Descriptor Specifications SIP Content Type Specifications SIP Sequencing Constraint Specification Formal Specifications Negotiated Prior to Detailed Packaging

13. PAIS – 16-18 January 2007 13 Main changes: overview Formal Definition Phase Transfer Phase Create Descriptor Specifications SIP Creation SIP Transfer SIP Reception and Validation by the Archive Archive responsability Producer responsability Producer and Archive responsibility Digital objects Time Formal Agreement * the process of data creation is not part of the standard. The data could exist at the beginning, or be created during the Process. * Descriptors SIPs Create SIP Content Type Specifications Create SIP Sequencing Constraints Specifications

14. PAIS – 16-18 January 2007 14 Main changes: Descriptors ■2 Descriptors:  Transfer Object Descriptor (TOD): set of characteristics describing a Transfer Object.  Collection Descriptor (no associated Data) : describes a collection of one or more Transfer Object Types or other collections.  gives a coherent view of the Data to be transferred. Descriptors are organized in parent-child relationships. ■Relationships between Transfer Objects, or Data Objects, or Transfer Objects and Data Objects, are described by ‘associations’ inside the Descriptor. ■General structure for the TOD?

15. PAIS – 16-18 January 2007 15 Main changes: Descriptor ■General structure

16. PAIS – 16-18 January 2007 16 Main changes: Descriptor ■Identification

17. PAIS – 16-18 January 2007 17 Main changes: Descriptor ■Description

18. PAIS – 16-18 January 2007 18 Main changes: Descriptor ■Relations

19. PAIS – 16-18 January 2007 19 ■Content Main changes

20. PAIS – 16-18 January 2007 20 Items pending: PAIS red book version ■Validate the final overview diagrams. ■Descriptors: open points (formats, …). ■Many exchanges between the teams have raised points that should be discussed this week.

21. PAIS – 16-18 January 2007 21 Items pending : SIP  SIP structure: suppress the attributes SIP Type location, Descriptor location?  SIP implementation using XFDU ■This part should be discussed along with the current version of the XFDU standard and the recently sent document:  To create a generic SIP Model to take into account the specificities of the PAIS, while complying with XFDU (possibility of using the same tools). This Model will be applied to all SIPs and all projects.

22. PAIS – 16-18 January 2007 22 Items pending: PAIS tutorial ■Create and structure a document ■Extract from the previous version explanations and examples as a basis ■Complete it to have an homogeneous tutorial (one practical example, …)

23. PAIS – 16-18 January 2007 23 Items pending: tool for the POT design and transfer follow up ■ A prototype has already been developped according to the previous standard version ■ Important to validate the generic Descriptors and have a new PAIS Version before working on the tool ■Many new things have to be taken into account for a new prototype (new Descriptor, SIP structure)

24. PAIS – 16-18 January 2007 24 We need ■ New version for a PAIS draft Red Book ■ New prototype ■ First version of a tutorial ■To be discussed: actions and planning

25. PAIS – 16-18 January 2007 25