Issue

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