1. COAR Global Metadata Store Scenario for an offer and its implications

2. Disclaimer 2 http://gisgis.blog.is/blog/brattur/month/2008/10/ This objective can only be understood if you become a Marsian and then look at the earthlings‘ world of Open Access Repositories!

3. Why would it be necessary? Vision: rich public and economical service layer based on OAR-share in Open Information Commons –examples: global indexing, search, statistics, mash-ups, alerting, recommender, re-use, mining, enrichment … OAR vision challenged by distributed nature –requires each service provider overcome heterogeneity, laboriously identify entities, aggregate metadata COAR Global Metadata Store could be a cure –reference object registry of OA-objects free for everybody to use (or at least non-commercial use) Benefits / Success Factors: increased visibility and re-use of local repository content > traffic, inlinks etc., Open Inf. Commons > # of service providers

4. Practical grounds & Other Examples Metadata Stores already there –OAIster, BASE, DRIVER, ANDS … BUT: too laborious for COAR –Rather start from object IDs, radically web- based – URIs, namespaces, authority files –Then build raw metadata store for re-use Other examples may exist >> CrossRef? –But none is global, open and free and operated specifically for OARs 4

5. The most simple thing possible Object Registry –Registry service, stable identifiers etc. Harvesting and Open Store –COAR members repository content > no indexing Machine interfaces for service providers –e.g. http, SFTP, RSYNC, OAI-PMH etc. > no GUIs Updates plus support for repository managers & service providers technicians on work-days 24/7 deployment at two sites –As authoritative service (SLA) ~ 5000€/month –Best effort may be for free

6. Implications: „With or Without“ Without COAR Global Metadata Store + No hassle with cost and service quality – No global authoritative starting point for service providers in distributed and possibly fragmented and complex landscape With COAR Global Metadata Store + first step to “open information commons”, symbolic commitment to the proliferation of a service landscape, operating “hard assets” desired – Introduces technology and complexity to COAR