Steps to take before Rollout

[steps in brackets] -> optional/not required before rollout, but connected.

• Updating of local institutional repositories
  • All handled locally, at the moment tag local records as "INSPIRE-mastered" which implies hands off for local editing "go to inspire instead"
  • [Longer term this might be replaced by bidirectional harvesting based on harvesting of different elements]
  • [even longer should be like Git....]
  • tag specific CERN rept series for inclusion (manual)
  • [later can be done by bibmatch]
• Create "beta" version of inspire that is regularly updated from SPIRES, and available for beta testing.
  • First step is to look at the XSLT/data model and make sure it is correct...should be done this week
    • Author ID imported correctly (desy author)
    • [later look at external author IDs]
    • License/Copyright in the metadata
    • Collections (this week)
  • Fulltext
    • From arXiv at least for indexing
    • samples of older docs as well as inst. docs.
    • harvesting workflow includes configuration for storage and display of fulltext
  • Then create updater scripts
  • Also do the final formatting on the beta (remove stars/commenting)
• Mirroring - everyone hosts "dark" mirror except for live server.
• [Accounts - rollout]
  • [Lightweight accounts no association with authors]
    • [ask Tibor about mixed login... can we mix kerb (i.e. apache?) with lw self-service or is this still a long way off. ]
  • [use emails as accounts and then, at a later date, match these emails to hepnames to suggest papers to claim]
  • [use kca and other lab infrastructures, that can also claim previously registered accounts]
  • [make sure that invenio confirms email via pinging the registered address]
• Inputting Tools
• Hepnames
  • Email privacy
  • [3 levels verification/contact/display (merged if account displayed only to logged in users) (and unknown/old.email) need all 3]

-- TravisBrooks - 2009-09-28