= Minutes to Meeting in Nijmegen 2010-09-16 =

== Relation Registry ==
Status: 
 * first prototype running
 * minimal test data (linking dublincore to isocat)
 * simple query REST-interface

Next Steps:
 * publish the interface

== Component Registry ==
Status:
 * stable, used
 * +10 users / 42 profiles/ 145 components

Next Steps:
 * view all anonymously
 * shibbolize connection to isocat
 * ways of curating - further constrain GenComp.XSD
 * think of versioning, trying to prevent usage of deprecated component-versions
   * automatically link to newest version of components, when editing a profile
   * provide scripts for transforming instances to newest version

== Abril ==
 * referencing external schemas - stay self-contained, except for xml:lang, employ OASIS catalog file

== Data ==

!ToDo:
 * diversify the dataset / group the data by providers (as root-level collections):
  MPI-IMDI::
   restrict the basic test-dataset to only MPI resources
   -> create new snapshot - without OLAC, provide the one root-collection CMD-record
  OLAC::
   make OLAC separate dataset
  CLARIN-EU-Inventory::
   data inputted by clarin-members on the homepage
  aac-test-corpus::
   Test dataset from Vienna

== MDRepository ==

!ToDo:
 * accept simple search - just a string query, multiple terms are treated conjunctive, phrase is in quotes
 * activate the test suite - collect the queries

== MDService/Browser ==
Status (new functionality):
 * view of the profiles with usage statistics
 * dynamic columns
 * load/save queryset/query

Next steps:
 * allow simple search
 * linkup with VCR:
  * publish the stored queries as intensional-VCs
  * collect md-records and publish them as extensional-VCs (needs some bookmarking)
 * allow to search via DatCats ([wiki:SemanticMapping#SearchinginDatacategories description of the procedure])
   * link up with RR
 * polish the layout (logo!)

== Other ==
 * [CmdiVirtualLanguageObservatory VLO2.0] -  a new more stable (solr-based) version to have something nice to show fast. [[BR]]
   (at least) start independent of MDRepository/MDService as they are slower, more experimental tracks, although an integration seems sensible later (MDService/MDRepository harvesting, doing some preprocessing, feeding the data into solr) 
 * Produce own public page for CMDI as a common entry-point presenting all the software-components (REST- and User-interfaces)
 * Push the standardisation of Component-Specification (ISO-Meeting Berlin)
 * Think of establishing a more stable organisation - an architectural board
 * Define a (one or multiple) ''WP5-Taxonomy''-profile, only to be used on the exploitation side (''searchByProfile'')