[Iccrg] philadelphia meeting minutes

[Eddy, Wesley M. (GRC-RCN0)[VZ]]

>-----Original Message-----
>From: iccrg-bounces at cs.ucl.ac.uk 
>[mailto:iccrg-bounces at cs.ucl.ac.uk] On Behalf Of Michael Welzl
>Sent: Wednesday, March 26, 2008 1:41 PM
>
>> 1. "Consensus" of the ICCRG is not entirely clear.  Does it 
>mean that 
>> everyone *subscribed* to the list agrees, or everyone who *posts* to 
>> the list agrees, or that those who attend meetings (Manchester?
>> Philadelphia?) agree?


The IETF usually talks about "rough consensus" rather than "consensus",
in my experience :).


>My god, this is the one big IETF mystery that I've been 
>wondering about all these years. I'll be happy to see the 
>explanation...


This was raised within the context of how we conduct safety
reviews of transport protocols (in general), and what our
"exit strategy" is for finishing reviews once started.  For
CTCP, it seems that we've hit a point where we have basic
agreement both in the meetings and on the list, so that lets
us put the question of establishing consensus off for a little
bit at least ... however, it becomes important again when
discussing CUBIC and then future proposals that are brought
here.  For CTCP, it seems to me like most reviewers arrived
at similar conclusions and didn't dispute the conclusions of
others, so we did not have difficulty obtaining "rough
consensus".

*I believe*, that if at some point it seems like it will be
difficult and time-consuming to obtain consensus, then we have
the leeway to issue a review that says there wasn't consensus
but that N reviewers backed by some list of material (papers,
emails, etc.) felt a proposal was safe for a stated level of
experimentation, and that M reviewers backed by some other
list of material felt that it was only safe for a lower level
of experimentation.  I may be wrong.