PIM-WG:

1. Status updates
2. draft-ietf-pim-drlb-03
3. Experimental/private join attribute types
4. draft-zzhang-l3vpn-mvpn-global-tble-mcast-03

1. Status updates

  draft-ietf-pim-explicit-tracking
   Submitted to IESG, wants some changes.
   Author to submit new draft and WG to review.

  draft-ietf-pim-explicit-rpf-vector
   Requested publishing. Reviewed by AD, waiting for new  version.

  draft-ietf-pim-rfc4601bis
   We should have done WGLC for this. WGLC right after this meeting ?
  
  draft-ietf-pim-drlb
   New version to be presented.

  draft-ietf-pim-hiarchicaljoinattr
   New v just submitted. Bascially reverted back to earlier version,
   removing some changes.

  draft-ietf-pim-igmp-mld-wireless-mobile
   No progress for this for this IETF.

  draft-ietf-pim-source-discovery-bsr
   No progress for this for this IETF.

  draf-arango-pim-join-attributes-for-lisp
   Adopted, but can not be published until some LISP wg. milestones completed.
   This draft does not depend on those milestones though, so review of
   this draft within PIM can proceed, but it can not progress through IETF
   before LISP WG reaches those other milestones.

  draft-sami-pim-ng
   Author sent emil to pim-wg. Wants the wg to consider the document.
   Please have a look and send comments to the author and the list ?

2. draft-ietf-pim-drlb

  Slides presented.

  Changes:
   DR election procedures remain unchanged
   A router announces hask masks in new Hello Option TLV
   to indicate its capability
   Hash masks include RP,Group,Source
   Modulo hashing is specified
   More can be added if necessary
   All candidate GDR must have the same DR priority as the DR
   DR announces the list of candidate GDRs and the hash masks to be used
   on this LAN.

   Creating forwarding states:
     upon receiving IGMP reports, a candidate GDR runs a hash algorithm to
     determine if it is the FDR
     If it is, it will do the forwarding
   GDR Assert
     Used to reduce packet loss during GDR state change
     A GDR becoming non-GDR MAY choose not to remove the oif immediately
     This will lead to both routers forwrding
     GDR asserts using normal metric
     non-gdr asserts using PIM_ASSERT_INIFITY-1
  
  Draft status:
     last updated at vancouver ietf88, changes proposed:
     introduced algorithm field in DRLBC option to allow future additiona
     algorithms
     specify modulo in the default to achieve load balancing
     Ip "interface ID" option present in Hello, which includes the Router-ID,
     then the router-ID part is used instead of the actual interface IP addrss.
  
  Other changes in -03
    add 32-bit mask in modulo hash to simplify the calculation for IPv6.

    adjust assert metric to pim_assert_infinity-1
    suggest drlbc and drlbgdr type 34 (0x22) and 35 (0x232)
  
  Next steps: Second WG chair (Mike) to chime in on decision of
  running WGLC in WG for this draft.

3. Experimental/private join attributes
 
   Discussion with ADs about the ability to add attributes to PIM
   easier.

   Currently no experimental attributes.
     64 code points, used 4 so far
       explicit RPF and LISP join attributes use 3 more.
   Allocation based on IETF review
     RFC needed (i believe), but may get early allocation when document is
     stable.
   
   fairly strict, should we introduce experimental ?

   Slide - how do experimental code points work ?
     RFC3692 recommends having at least one value for experiments
       for experiments only, shot not be used in products, different
       experiments are likely to use the same value, hence not for
       general deployments.

   Slide - privte code points
     Set aside some attributes for mor private use
      Less strict, eg: with or without documentation
      registry needed to avoid conflicts
      expand the type space
     
     or do we believe that attribtue should be reviewed
     if we allow private code points, would anyone ask for code points
     that require review ?
    
     Toerless: Does not like experimental code points. Private code 
     points are good. making it easy to get private code points is good,
     providing documentation for privte code points without necessarily
     arguing them is good.

     Another audience member chiming in and saying he agrees.
  
  Encoded-Source Address Encoding Type Field Registry

  draft-atwood-pim-reserve-exp-00
    New draft submitted this Monday
    Proposes 2 expeimental attribute code points
    and 4 experimental encoding types

    Adrial Farrel (AD): Some of this depends on the scope of the experiment.
    If you want to try to do a mixed deployment, not crash somebody in the
    middle of it. This is the significant purpose of this.

    Toerless: Good reason, but fail to remember across various protocols a
    single good example where this was well used.
  
4. draft-zzhang-l3vpn-mvpn-global-tble-mcast-03

   repeatng presentatin from MBoned, ca. 30% of the room was not in MBoned.
   Not repeating text written there, see notes from mboned WG.

   Thomas Morin: L3VPN co-chair. Asking for adoption into L3VPN,
   going to send heads-up to Mboned/PIM, and ask for support from PIM-WG chair.