ESciDoc Committer Meeting 2010-06-29

From MPDLMediaWiki
Revision as of 08:02, 2 July 2010 by Mra (talk | contribs) (→‎Outcome)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Date: 29.06.2010 Start time: 14:30

Location: Karlsruhe, München phone: +49-89-38602-223 VidCo-ISDN: 08938602595

Participants MPDL: Natasa Bulatovic, Wilhelm Frank

Participants FIZ: Steffen Wagner, Frank Schwichtenberg, Michael Hoppe, André Schenk

Previous committer meeting

Next committer meetings

Topics[edit]

Release 1.2[edit]

Outcome[edit]

  • Migration 1.1.3-> 1.2
    • two migration steps that transform the foXMLs and will be executed if the Migration tool is run
    • OAI-DC transformation has to be done locally
    • OAI-Server has to be stopped (in any case it is not used now at MPDL due to the bug in 1.1.3)
    • do stylesheet transformation on the FoXMLs after core-service migration is done
    • in future when xslt for oai-pmh is changed, oai-pmh is done by itself
  • Wildcard matching is not implemented in 1.2
    • like operators are slow
    • check possibility for posix operators
    • if needed to be extended: depends on admin search schedule
  • Content Relations
    • further discussion is needed things to be clarified better. For now now change in the implementation.
    • the question is if a Content Relation object has to be 1:1 with a statement in the triple store
    • MPDL to check on possible use cases
    • potential solution for 1:M relation would be item with specific content model and usage of "links" e.g. simple content-relations from one to another item
    • the decision on deprecating the simple content-relations from Item Handler is discarded. FIZ team decided and communicated that the simple content-relations (in item.xml) will stay and the possibility to create them will also stay
    • in future there will be single interface for querying of all content relations (simple links or objects with metadata)
    • challenging for the documentation (how to describe the difference)

New Features[edit]

  • Admin Search: discuss new questions
  • PidCacheManager: Update
  • PubMan: Feature requests from FIZ side for next version?
  • Batch Updates: rambling thoughts
  • New roles (MPDL instances):
    • escidoc:role-cone-closed-vocabulary-editor
    • escidoc:role-cone-open-vocabulary-editor
      • Since version 6.1 of PubMan, CoNE vocabularies can be edited by the end-users. There are 2 types of Vocabularies (opened, closed). All users who have respective privileges can edit any open/closed vocabulary. At the moment, the authorization is done purely by CoNE internal logic. In future, privileges will be assigned directly on vocabularies. This improvement had been discussed on some previous Commiter meetings. To allow for creating external roles/grants (i.e. grants on objects not stored in eSciDoc repository).
  • eSciDoc_Admin_Roles
  • New roles: escidoc:role-context-administrator, escidoc:role-context-modifier, escidoc:role-user-account-administrator, escidoc:role-user-account-inspector
  • Modified roles: see Talk:ESciDoc_Admin_Roles

Outcome[edit]

  • Pubman improvements
    • Configurable parts in PubMan shall be better externalized, at the moment the .ear file has to be edited
    • instructions for how to deal with the validation, how to set-up and edit the validation schemas provided with PubMan
    • instructions for customizing the metadata
    • cmodel with pubman installer has to be created rather than required from the core service
  • Admin search
    • in first phase the admin search will be implemented as proposed in Colab
    • hierarchical permissions are problematic with respect to time to reindex in case of deeper container hierarchy with many members -> for start they will be implemented as proposed in Colab, if there are further issues these will be resolved further
    • With 1.2 one can change the stylesheet that states which fields are indexed
    • With the next version one can additionally configure:
      • new indexes
      • which resource types to index
      • when to index (only released or also pending...)
      • when to remove from the index (e.g., withdrawn)
      • which analyzer to use
      • how to Highlight Search-Result
      • configure per index if indexing is done asynchronously or not
      • what to return in search hit (e.g., full resource representation or only ObjectId)

Miscelaneous[edit]

  • Developer Workshop (September 22/23, 2010 in Karlsruhe): Topics?