On 11/06/2015 10:48, Joel M. Halpern wrote:
>
>
> On 6/10/15 5:58 PM, Michael StJohns wrote:
>> Let me try this again.
>>
>> 1) Is my description of the IETF process reasonably close to reality?
>> E.g. does the consensus process contribute to "Standardization by
>> Combat"?
>
> Sometimes. Not Always.
I think, specifically, that our rough consensus process is not conducive
to compromise between alternatives, whereas "voting (consensus preferred)"
does tend to lead to compromises. By "compromises" I mean things like
the ATM cell's payload size being the arithmetic mean of 32 and 64,
or the OSI Network Layer having two incompatible protocols. So it may
be a feature, not a bug, that our process discourages compromise.
When we have to choose between distinct alternatives, the next point
applies:
>>
>> 2) If my description is not exactly correct (or always correct), how
>> does reality differ from this description?
>
> From where I sit, the difference lies in how the chairs manage the process when things get rough.
Exactly. And things can get rough quickly and unexpectedly.
>>
>> 3) If my description is correct, can the process be changed without
>> changing the fundamental nature of the IETF?
>
> There may well be ways to improve the process. Pete Resnick's efforts to clarify what we mean by rough consensus are probably
> an (unfortunately necessary) step towards such improvements.
True, but things can get very heated way before there is any real
question of a consensus call. This isn't at all easy for WG Chairs
to control, especially between meetings.
Brian
>
> Yours,
> Joel
>
>>
>> A few comments in line.
>>
>>
>>
>> At 05:41 PM 6/10/2015, Eric Gray wrote:
>>> The biggest problem with this approach is that it tends to work
>>> more for people who are good at winning arguments, using whatever
>>> tactics they choose, over those who are right - on those occasions
>>> when the two are not the same.
>>
>>
>> So is this a characteristic of the IETF or not? Never, Sometimes,
>> Always?
>>
>> In any event, it's not about who's right, its about what's useful to
>> solve the problem. Which causes problems when there are many ways to
>> solve the problem, each reasonable, and each supported by its own
>> choir.
>>
>>
>>> Not all bright people are able to overcome an innate introversion
>>> to the extent that is required to be successful in a shouting
>>> match.
>>
>>
>> Counter point: Not all bright people are able to understand that
>> they are not always the fount of all wisdom and that shouting out
>> their brilliance will not necessarily accomplish what they want to
>> accomplish. However, the current model does deal with this set of
>> behavior reasonably well.
>>
>>
>>
>>> And some of the brightest would rather see us flounder as a group
>>> while they take their arguments elsewhere.
>>
>> This sounds suspiciously like "they'll take their toys and go play
>> somewhere else"? Which isn't really good behavior for adults IMHO.
>>
>> Mike
>>
>>
>>
>>> Just a thought... -- Eric
>>>
>>> -----Original Message----- From: ietf
>>> [mailto:ietf-bounces@xxxxxxxx] On Behalf Of Dave Crocker Sent:
>>> Wednesday, June 10, 2015 5:26 PM To: Michael StJohns; IETF
>>> Discussion Mailing List Subject: Re: discussion style and respect
>>>
>>> On 6/10/2015 9:40 PM, Michael StJohns wrote:
>>>> Through "consensus", we include things that are strongly
>>>> presented, vigorously defended, said by people with gravitas
>>>> applicable to the technology[, technically good], and not shouted
>>>> down. It may be that the style of interaction that you're
>>>> complaining about is more related to the "consensus" process than
>>>> to any other element. If may be that if you want to change the
>>>> confrontational style, you're going to have to change the way
>>>> things become standards.
>>>
>>>
>>> In spite of formal voting, some other standards groups either
>>> explicitly or implicitly use a unanimity (not 'rough) consensus
>>> model. Still, they do not suffer anything approaching quantity of
>>> rude and disrespectful behavior that we tolerate and, arguably,
>>> condone.
>>>
>>> Adult, respectful behavior occurs when it is required. We don't
>>> require it.
>>>
>>> Not really.
>>>
>>> d/
>>>
>>> ps. Periodic, generic -- albiet heartfelt -- pleas for better
>>> behavior might be necessary, but they have had no effect -- ever --
>>> in almost 30 years.
>>>
>>> -- Dave Crocker Brandenburg InternetWorking bbiw.net
>>
>>
>>
>
>