...

DSpace architecture white paper by RobertTansley - dspace-architecture-whitepaper.pdf\

...

This page is intended for those of you doing DSpace work to provide a link to wiki pages you may have created to support your development work or to document your procedures.
Typically code will be stored in the "sandbox" area of the source tree.

• Dynamic Browse Prototype - RichardJones
• OAI Sets Generalisation - RichardJones
• Persistent Identifiers (requirements) - JamesRutherford
• DAO Prototype - JamesRutherford
• About Data Formats – Larry Stone
• BitstreamFormat Renovation – Larry Stone
• Embargos - Elliot Metsger
• Hiding Collections/Communities - Elliot Metsger
• Modular Usage Statistics - Mark H. Wood
• Tagging - christophe_dupriez
• Simple Archive Format - Packager - Peter Dietz

Architecture

Work towards the DSpace 2.0 architecture: DSpace 2.0 Original Proposal

Storage refactoring

• AssetStore - standards-based AIP storage layer for easier preservation
• AipPrototype - (OBSOLETE - AIP Backup and Restore has replaced this). Please use prototype AIP implementation, above the asset store layer, for use on DSpace 1.4/1.5. Emphasis is on preservation and custody transfer of archived objects.
• PluggableStorage - allow easier integration of different asset store backends. Prototypes within.
• DAO Prototype - removes storage layer depedendent code from core classes. Should make it easier to port DSpace to other database platforms.

Modularity

• ModularityMechanism - Rather old musings
• PluginManager - Simple first steps
• AddOnMechanism - Defining a mechanism for build-time add-ons

User interface framework

• UI Framework
• Internationalization Support (I18nSupport) - Internationalisation progress

Other aspects

• EventMechanism - Adding event mechanism to the data model
• EventSystemPrototype - prototype implementation (for DSpace 1.5) of an event driver
• ScalabilityIssues1.4
• CleanupTasks - General tasks to ensure the architecture/code is robust and consistent
• PersistentIdentifiers - making DSpace less dependent on the Handle system.
• BitstreamFormat Renovation - new design and implementation of data format technical metadata for Bitstreams, to better support digital preservation.

  ---

Features/functionality

• DSpaceStatistics
• PersistentIdentifiers - Making identifiers 'pluggable' (Handles, ARKs...)
• NetworkInterfaces - e.g. Web Services, WebDAV etc.
• HistorySystem
• Item Versioning Support - approaches to supporting multiple versions over time (i.e. editions, as opposed to multiple formats of the same version)
• BitstreamRelationships - proposal to encode relationships between bitstreams explicitly in the data model.
• Citation Cover Page - generating a Citation Cover Page to PDF's downloaded from DSpace on-the-fly

Metadata

• MetadataSupport for non-Dublin Core schemas
• XmlNamespaces
• DSpaceIntermediateMetadata - XML format for metadata tables for XSLT transforms
• CrosswalkPlugins
• PackagerPlugins
• MetadataDiscovery
• Attaching "policy" metadata to objects, SIPs (experimental)
• DSpaceMETSSIPProfile (for use with the Packager Plugin, via the Lightweight Network Interface and other package-based applications)
• DSpaceMETSAIPProfile (used by AIP Prototype)
• ItemBatchUpdate

Security

• AuthorisationSystem
• StackableAuthenticationMethods

Preservation

• DigitalPreservationToolsAndStrategies
• AIP Backup and Restore - Based on the initial AipPrototype implementation (by Larry Stone), but only details with exporting/importing Archival Information Packages which describe the entire content hierarchy (Community/Collection/Item) within DSpace. Emphasis is on creating a better "backup" and restore solution, where an entire DSpace could be restored from a directory of these AIPs. [Released in DSpace 1.7.0]
  • AipBackupRestoreDiscussion - Discussion about the AIP Backup and Restore feature released in 1.7

Search

• SortingSearchResults

Installation / Updates

• Installer Prototype - Prototype for a new "easy" installer, which will better automate the Installation/Upgrade processes for DSpace.

Big Data

• Ingesting Large Bitstreams - uploading big data sets via HTTP

...

• Half-developed: Diagnostics servlet. A servlet that goes around and kicks the apps's tyres a bit on startup or on request. At the moment this just checks that the lucene index is present and unlocked. I'm hoping to extend this to checking that the db is present and correct, and maybe to check that the db and filesystem are in synch. No schema changes. Ojd20

• Largely conceptual, some code snippets exist: Upgraded submission system (SubmissionSystem) - this is quite a big job so perhaps not for 1.3 but we'll see. RichardJones

• Conceptual, shouldn't be too hard: Upgraded workflow system (WorkflowSystem) - not too big a job I don't think. Hardest part will almost certainly be the interface, as the options could be a little convoluted. RichardJones

• Hierarchical collections as well as communities. Not implemented.