[15:48:16] toby moncaster joins the room
[15:56:08] toby moncaster leaves the room
[16:00:14] toby moncaster joins the room
[16:03:12] slblake joins the room
[16:03:20] <slblake> Howdy
[16:03:25] <toby moncaster> hi
[16:03:50] <toby moncaster> I'm also running remote. I'm hoping someone in the room will also log into jabber...
[16:05:11] Lars joins the room
[16:06:19] <slblake> Hi Lars
[16:06:24] <Lars> hi
[16:19:30] <toby moncaster> Can I make a point - it seems as if Daisuke is trying to do new research. I thought IETF was meant to be for proposals that have already been researched and in my view the current marking behaviours have been researched to death...
[16:21:57] <slblake> We allowed the option for publishing experimental encodings.
[16:22:26] <slblake> I agree that on this topic we have long passed the point of diminishing returns.
[16:22:51] <toby moncaster> I'm perfectly happy for their to be experiemntal markings in the future...
[16:56:24] <Lars> i might have misunderstood, but it did sound like we wanted to change the 2-state marking, no?
[16:56:38] <Lars> s/we/he/
[16:57:05] <slblake> I didn't catch it exactly
[16:57:38] <toby moncaster> I wasn't clear eitehr
[16:58:57] <toby moncaster> ecodings are in my baseline slideset
[16:59:37] <toby moncaster> PDF
[17:05:25] <toby moncaster> packet specific dual marking. requires you to know which Not marked state a packet originally carried before being marked. E.g. only works with a RSVP trype signalling
[17:09:18] <toby moncaster> I would expect us to make them somehow related.
[17:10:40] <toby moncaster> or even worse Friday evening!
[17:11:49] <toby moncaster> Could the experimental one come from pool 2 (if that's the one that can be used in future for standards as well as local use?)
[17:13:59] <slblake> IANA shouldn't allocate out of pool 2 until pool 1 is depleted.  So there is no advantage to doing an experiment with pool 2 values; you will likely still have to change if your scheme is standardized.
[17:15:36] <slblake> All the boxes I have worked on allow you to configure the mapping between DSCP and PHB.
[17:15:41] <slblake> As intended
[17:15:57] <slblake> And also allow you to specify specific DSCP values in ACLs
[17:17:17] <Lars> steve: want this to be relayed?
[17:17:27] <Lars> to the mike?
[17:18:43] <slblake> It would be worthwhile for people doing PCN experiments to notify the community (PCN list + TSVAREA?) which EXP/LU codepoints they were planning to use, just for the convenience of others who might want to participate in multiple experiments.  We (the PCN chairs) could maintain an informal list of EXP/LU DSCPs  in use (that we have been notified about, of course)
[17:18:51] <slblake> ok
[17:20:25] <toby moncaster> pool 1 standards, pool 2 is LU until pool 1 is exhausted
[17:21:03] <Lars> bob says: steve, please post this to the list
[17:21:08] <slblake> will do
[17:21:37] <slblake> FYI, from RFC 2474:
[17:22:07] <slblake>    Pool        Codepoint space          Assignment Policy
   ----          ---------------                 -----------------

    1             xxxxx0                            Standards Action
    2             xxxx11                            EXP/LU
    3             xxxx01                           EXP/LU (*)

   (*) may be utilized for future Standards Action allocations as
       necessary

[17:22:20] <toby moncaster> are Bobs slides online anywhere?
[17:22:45] <Lars> they are part of the tsvarea slide set from yesterday
[17:22:49] <Lars> or on his private page
[17:22:53] <slblake> so it is actually pool 3 that may get  taken over for standardized codepoints by IANA in the future.
[17:23:25] <toby moncaster> http://www.cs.ucl.ac.uk/staff/bbriscoe/presents/0903ietf/0903tsvwg-ecn-tunnel.pdf
[17:39:32] <slblake> Are his slides on the meeting materials page?
[17:39:41] <slblake> Kwok's, that is.
[17:39:47] <toby moncaster> not that I can find
[17:40:05] <Lars> no
[17:40:14] <Lars> he didnt get them to scott in time
[17:40:21] <slblake> ok
[17:41:27] <slblake> just curious, what is the attendance like this week?  Normally I would expect it to be high due to the location, but with the economy I expect some decrease.
[17:41:37] <Lars> ~20
[17:41:59] <slblake> I meant IETF attendance, but I am also interested in PCN attendance (which is probably reduced due to the Friday timeslot anyway)
[17:42:35] <Lars> ah
[17:42:37] <Lars> ~1200
[17:42:39] <Lars> unchanged
[17:42:44] <slblake> that's not too bad
[17:42:50] <Lars> but about ~150 more from CA then usual
[17:42:54] <toby moncaster> I had a quick look at the registered attendance which was 1300+
[17:43:16] <Lars> so increased local attendance is compensating for reductions in global attendance
[17:43:29] <Lars> sweden and japan will be interesting ietfs
[17:43:41] <slblake> yep
[17:44:16] <slblake> It's fairly easy to get to Stockholm from Europe, although the hotels are expensive.  
[17:44:19] <toby moncaster> Inc ~100 Cisco but only ~30 Juniper despite them being sponsors
[17:44:41] <toby moncaster> Japan will be really low attnedance I reckon as both US and Europe will struggle to travel
[17:44:56] <slblake> nothing is more entertaining than watching 3 Cisco people arguing with each other in a wg meeting.  :)
[17:45:20] <toby moncaster> which reminds me - I meant to ask how the LISP BoF went?
[17:45:52] <slblake> I listened to most of it.  My perception is that a wg is a done deal.
[17:46:14] Lars leaves the room
[17:47:46] <slblake> Juniper doesn't seem to be very interested in LISP (nor AlcaLu nor Redback), so the LISPers will have an uphill battle.
[17:48:37] <slblake> Gotta run.  Hope to see you in Stockholm.
[17:49:08] slblake leaves the room
[20:12:02] toby moncaster leaves the room