Hums and mailing lists (Re: Problem statement draft comments)

[Harald Tveit Alvestrand]

--On 13. mars 2003 17:34 -0500 Thomas Narten <narten at us.ibm.com> wrote:

> And of course, the final outcome needs to be ratified on the mailing
> list. But, if the hum is done right, by definition there shouldn't be
> major push back on the mailing list. I.e, if the hum agrees on X, but
> there is a lot of pushback on the mailing list, that raises basic
> questions about whether enough of the right people were in the room
> for the hum to go on a particular direction. Or that the issues that
> are coming up on the mailing list were adequately considered before
> the hum took place.

there was one rather famous example (in the security area, I believe) when 
the WG had an "unsolvable" problem in the mailing list, had a physical 
meeting with one near-unanimous hum (the meeting lasted for all of 15 
minutes), and then took the resolution back to the mailing list - which 
exploded, since the main opponent to the proposed solution wasn't at that 
IETF, and he managed to convince quite a few people that he was right and 
the hum was wrong.
I'm sorry that I don't remember the WG or the year, or the final outcome; I 
wasn't directly implicated, but think it must have been in the mid-90s.

(this was also one of the events that led us to consider introducing 1-hour 
slots on the agenda; for some meetings, we know beforehand that 2 hours are 
not needed...)

                        Harald