View source

A list of issues about which decisions need to be made.  See also [[Work packages]].

* [[/Form of URIs]] (host name(s), paths, etc) (see also [[URI Format Considerations]])
** [[/Versioning]]
** [[/Databank names or identifiers]]
** [[/Case and canonicalization]]
** [[/Provider partitions‎]]
** [[/Provider partitions/Biased unqualified]]
** [[/Provider partitions/Normalize taxon]]
** [[/Derivative resources]]
** [[/Purl extensions needed]]
** [[/OCLC noise tradeoff]] (resolved?)
* [[/One ID per record‎]]
* [[/Requirements for being a server]]
* [[/Server online failure]]
* [[/What metadata to record]]
** [[/How and whether to indicate what records are about]]
* [[/Knowing when we've requested a record that doesn't exist]] (404)
* [[/Addressing target databank instability]]
* [[/What distributions]]
* [[/Where to host]]

* Form of RDF (what triples)
* Form of HTML generated from RDF via XSLT
* Ontology to be used (by the generated RDF)
* Media type (application/xml vs. application/rdf+xml vs. application/xhtml+xml)
* Hash versus slash
* URI format upgrades
* Business model (not a technical issue, put elsewhere in wiki)

* Primary form in which curated idspace information is stored
* User interface for updating idspace information (curation)

-----
== (From an older issues list) ==
* How to initialize the databank registry
* How to maintain registry of third-party value-added services
* Choice of software platform
* How to implement redundancy and failover
* Ways to access server information - e.g sparql, downloads of various forms, feeds