Version 5 (modified by 16 years ago) (diff) | ,
---|
Meeting between MPIWG group & eSciDoc team members, 16-17 October, Munich
MPIWG participants: R. Casties, M. Hyman, W. Schmidle, K. Thoden, J. Willenborg
eSciDoc architecture as of 2008-10-16
- three layers
- common functionality
- generalized and unified data model for all content resources
- content modeling of resources as a specialization instrument
- versioning
- PID
- currently a political problem; to be implemented really for faces etc. by year-end
- user management, authentication and authorization
- intermediate services (added functionality)
- duplicate detection
- image handling
- metadata handler
- validation of data
- retrieval/download statistics
- workflow management
- application services (can be integrated)
- depositing
- publishing
- quality assurance
- citation manager
- export manager
- [nowiki SearchAndOutput?]
- controlled vocabularies
- common functionality
- use of Fedora
- Fedora has its own versioning model
- versioning model is component-level
- no version numbers (data-based versioning)
- core services of eSciDoc implement versioning independently
- an eSciDoc item may be more than one Fedora object
- three stages
- pending
- submitted (some quality control)
- released (a published, valid object)
- after release an object cannot be deleted (to ensure persistence)
- interfaces
- SOAP and REST
Overall evaluation
- core eSciDoc services
- may be of some use, but developed functionality remains unclear
- scheduled release date for 1.0 is Q1 2009
- missing features
- e.g. notification mechanism (for letting us know that a relevant document has been created or modified)
- design features & flaws
- solutions (links below may break at any time, since they are to demo resources)
- Faces (MPI for Human Development)
- Virtueller Raum Reichsrecht (MPI for European History of Law)
- PubMan
- comments
- partner institutions for which a solution is being developed are "privileged"
- no public or stable repository exists yet
- future of project is unclear
- need to reevaluate situation in Q1 2009