- presenting agenda
  - No-one to present new HMAC draft
  - Ran: we will take draft into WG LC?
  - Chopps: yes.

- Les presenting EXTLSP update
   - version 2 update
       - resubmit for std track
       - conflict in metric of tlv22/23 clarifies conflict (i.e.: use tlv22)
- Les presenting GENAPP update
   - WG doc now.
   - updated for std track
   - applicability statement (routing vs non-routing info)
   - possible that we have to re-evaluated applicability compared to
     existing info carried in isis.
   - wg to decide which info goes where
   Chopps Q: any work in parallel with ospf ?
   Roy       A: we presented similar draft in ospf wg
   chopps: same decisions should be consistent in isis/ospf in terms of
           what goes where
   Q: how to decide existing rfc's ? are we going to impose to revise
      existing rfc's ? inventory base ? screening of existing rfcs ? just
      asking...
   Les A: news info is straight forward. for exiting tlv's it will be
          discussed in wg. no intention to force existing tlv to move.
   Chopps: moving target. we're not deciding anything now.

- Les presenting isis and BFD draft
   - use of bfd of isis
   - isis does not share forwarding fate with bfd (ipv4/6 and clns)
   - prevent isis adj. if bfd is not set.
   - see presentation slides and draft.
   - ask for WG document (was it before ?)
   - Q: what if bfd session takes time more than holdtime ?
     A: isis session goes down
     Q: is it good to bring down isis adj ? maybe it's good to bring isis
        up even if bfd is down and start the state machinery after bfd is up.
   chopps: if bfd never comes up you will never detect failure
   Les: we buying some time. should be sufficient to bfd to come up. if not
        then assumption is that you have forwarding failure
   Q: concern: impl. support bfd . another router has buggy softw. you
      bring down adj because of buggy soft.
   A-chopps: isn't this normal ? you have to pickup a timeout or you don't
             detect.
   les: if impl. has problem.... then you have to react to that....
   chopps: wg document before. any objection ?
   no objection, WG doc.

- George Swallow on bitvector draft
   - gather reaction on new idea...
   - see presentation slides/draft.
   George Q: feedback  /120 v6 assumption.
   A:    I don't think it's a major problem in doing that.
   Q: what if vector different ? what if abr summarization mask is not same?
   A: you do what we do today. most specific summary.
   Bruno Decraene: bgp NHT, does the drat address tunnel mode or hop-by-hop?
   Stefano rephrases: do you need a tunnel to next-hop or does it work with
               hopbyhop routing
   George: routing hop by hop and work well with tunnels.
   BD: should bgp use the bit vector ?
   G: bgp registers with isis to find out the next-hop based on bits
   BD: case of hop by bop. forw. loops if routers not compliant/supporting.
   SP: yes, but this is deployment issue. Full support is required before
       enabling it.
   G: the idea is also targeted for tunneling.
   Q: instability in one area will trigger generation of bitvector. churn
      is in routing plane not in fib
   Kireeti: if point is tunnel ldp vector ? isis is ip reachability.
       there are other tunnels (IP) but if you put it in isis it's for ip only.
       nht for ip if isis.

- doc status by chopps
   - require 2 independent implementations
   - non legacy requires implementation reports.
   - legacy not requiring implementation reports
   - see slide on rfc/draft numbers/names
   - rfcs completed
       - see slides
   Tony: mea culpa (3 times)
   - comments ?
   Kireeti: gmpls-isis std ? do you want ?
   A: it has to be because of referenced
       deployed:
   Kireeti: not implemented

END of Meeting