[db-wg] Draft minutes of RIPE 68 meeting

Dear WG,

I realised whilst going through the action points from last meeting that 
I hadn't sent out the draft minutes from RIPE 68. My deep apologies for 
this, I usually try to get these out immediately after the meeting.

Draft minutes attached.

Nigel

-------------- next part --------------
Draft Minutes for the DB-WG meeting at RIPE 68

A. Administrative Matters
   . Welcome
      Ulf Klieber is standing down for personal reasons
   . select scribe (Nigel Titley)
   . finalise agenda
      No additional agenda items
   . approval of minutes from previous WG meeting(s)
      Approved
   . review of action list
   AP66.4 Abandoned
   AP67.1 Complete (See below)
   AP67.2 Complete (See below)
   AP67.3 Complete (See below)
   AP67.4 Ongoing
   AP67.5 Ongoing
   AP67.6 Ongoing


B. Data Base Operational Update (Denis Walker, Ed Shryane, RIPE NCC)
   See presentation
      All issues are now tracked in github
         (although this was noted to be not necessarily useful to people who
         were not familiar with Github)
      Code base is open and in github
      All releases are now made to a release candidate before going to production
      The via: attribute has been added to autnum
      Member resources are covered by abuse-c:
      Roughly 33% of IPv4 and 44% of IPv6 PI space is covered by abuse-c:
      Single Sign On (SSO) has been introduced which will allow the introduction of more 
      features and services such as authenticated queries.
      Pending route authorisation gives a hierachical authorisation process.
      Documentation has been improved and is now linked to the version number
      Hot node in Stockholm is now up and running and accepting queries (load
      balanced with Amsterdam)
      aut-num now has a mandatory status ('ASSIGNED', 'LEGACY', 'OTHER') (automatically
      added)
      All inetnum legacy objects have been set to status: LEGACY
      sponsoring-org: has been added to end user objects

   Unresolved features 
      Deprecate "referral-by:" 
      New flag to request personal data. Should this go ahead?
      Only block person objects when limits reached (rather than everything)
      Internationalisation? What business/syntax rules should we apply?
      Replace changed: with created: and last-updated: attributed. Stats show that 27.4% 
      of changed: dates don't match the actual data.
      Improved dummification algorithm removing slightly less data and standardised

   Upcoming features
      Improved user experience
      Review of all error messages (does anyone parse them?)
      History query to show deleted data
         (it was noted that this potentially allowed the full history of an address
         block which one speaker thought may be a problem)
      Role object no longer considered personal data (remove personal data)
      Convert irt: object to role: object (and deprecate the irt: object)
      Change syntax of role: to match organisation: (remove limitations on name (for example))
      Allow change of person: and role: names (no legal issues)
      Authenticated queries will allow unlimited access to personal data (for example)
      Full re-implementation of auto cleanup including clusters of data
      Deprecating object editors by using business rules to lock down appropriate
      fixed object parts
      Server side as-set expansion (client performance improvement)


C. New DB-Software functionality
   - in test, scheduled for deployment (sponsoring LIR)
      See above. Sponsoring-org: is set to a dummy organisation at present in the RC
      database.

D. Removal of attributes
   - consensus on removing referral-by:, implementation
      Consensus has been reached. The NCC should remove this [AP68.1 [RIPE NCC]]
   - consensus on removing changed:, implementation
      Although there appears to have been consensus on the mailing list, RV objected and was asked
      to express this on the mailing list
   - status and implementation of last-modified - any open issues?
      Not as far as we have seen

E. 2012-07 implementation
   See above
      
F. Status update on IETF WEIRDS activities (Ed Shryane, RIPE NCC)
   See presentation
      WHOIS is being replaced by WEIRDS for a number of reasons including a limited data model, no internationalisation,
      no authentication, no redirection and insecure.
      WEIRDS covers names and numbers and uses the RDAP protocol over HTTP and HTTPS
      It covers a limited number of objects. Contact should be made with the WEIRDS-WG to get
      additional objects added.

G. Personalised Authorisation (Denis Walker, RIPE NCC)
   See presentation
      The basic issue is that people do not understand the concept of a maintainer.
      It would be better to group persons under a role associated with an org and use it
      as a maintainer.
      This integrates better into SSO. 

Y. Input from other Working Groups and/or Task Forces
      Review of Abuse Contact Stuff

Z. AOB
   - removal of geo-location functionality?
      No idea of who is likely to maintain this data. However there is now a crowd-sourced
      effort under way to try and maintain this data. Can the community please indicate if they
      are using it.
   - removal of white pages service?
      Is this now necessary? 
   - creation of a TF for a major review of the DB environment?
      We probably need a TF to push forward with proposals for the un-resolved issues above.
      This could be an interim meeting between now and the next RIPE meeting. Please see either
      of the WG chairs.