Re: limiting our set of cities

Christian Hopps <> Thu, 20 February 2020 20:19 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 6DE61120125 for <>; Thu, 20 Feb 2020 12:19:38 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.897
X-Spam-Status: No, score=-1.897 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_NONE=0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id YfexyUgUuvjm for <>; Thu, 20 Feb 2020 12:19:36 -0800 (PST)
Received: from ( []) by (Postfix) with ESMTP id EC5CB120112 for <>; Thu, 20 Feb 2020 12:19:35 -0800 (PST)
Received: from ( []) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by (Postfix) with ESMTPSA id 44EF960B76; Thu, 20 Feb 2020 20:19:35 +0000 (UTC)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 13.0 \(3608.\))
Subject: Re: limiting our set of cities
From: Christian Hopps <>
In-Reply-To: <>
Date: Thu, 20 Feb 2020 15:19:34 -0500
Cc: Christian Hopps <>,
Content-Transfer-Encoding: quoted-printable
Message-Id: <>
References: <> <> <17764.1582194882@dooku> <> <> <> <> <>
To: Keith Moore <>
X-Mailer: Apple Mail (2.3608.
Archived-At: <>
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: IETF-Discussion <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 20 Feb 2020 20:19:39 -0000

> On Feb 20, 2020, at 12:57 PM, Keith Moore <> wrote:
> On 2/20/20 9:13 AM, Christian Hopps wrote:
>>> On Feb 20, 2020, at 8:17 AM, Keith Moore <> wrote:
>>> On 2/20/20 7:46 AM, Christian Hopps wrote:
>>>> I think that we should pick the top 12-16 locations that participants are from, then for each destination prior to it being considered we calculate the travel PAIN (cost + time) for that set of participants.
>>> Why favor participants from large cities?   It's not like they're representative of the whole group.
>> By trying to make it easier for the most people, of course it's going to be helping areas with the most people. The point is to obtain a list of sites to measure travel cost and time from.
> I think that's statistically incorrect.   The likely effect of what you propose would be to artificially discourage participation by people not living in large cities.

It would only be statistically incorrect if most people didn't live in large metropolitan areas/cities. I seriously doubt the majority of IETF participants are rural.

FWIW *I* am rurally located. :)

I'm a plane ride away from (DTW/MSP/ORD) airports and I'm perfectly fine with the Detroit(AA)/Chicago/Minneapolis representing me with their pain. I'll always have that extra plane ride, and the rest of the trip will be minimized along with those "nearby" large city folks.

>>> Also, different participants have different ideas of pain.
>> I think it's reasonable to equate "painful" with travel time and cost.
> For me, pain includes those things but also includes things like the availability of food that is compatible with my diet.   And not just travel time but discomfort, which is nonlinearly related to travel time.

But what I've been talking to is the travel effort that is talked about in the mtgvenue documents/policy. Dietary restrictions are not a part of travel effort/pain. I think we need to avoid getting too abstract here until we figure out how to at least get the easy metrics optimized.

>> Do we really want people who love to travel and couldn't care less where we go to be diluting the "pain pool" for the measurement?
> I have yet to meet anyone who doesn't have a preference about where they travel.   But I think that letting everyone decide for themselves what is painful and whether the meeting location is an effective enough place to work to justify the trip, is much better than having one person unilaterally declare what the criteria should be.
>>> A fairer method would be to poll every participant about their preferences for future meeting cities, then for each meeting, pick N polled participants at random from those who have attended the last M meetings (locally or remotely), and select from the cities show up in their preference lists.
>> Ok as long as N is large and M is reasonable. But making N large is just going to skew to large metro areas being heavily represented anyway. *shrug*
> The difference is that the chance of considering a small city as a location would be in proportion to the number of attendees from small cities.

I can think of many reasons we don't want to meet in a small city, even if (and again I seriously doubt this) a majority of participants were from small cities. The point is many people to gather in one place. Large cities have already been optimized for exactly that condition.

> But I also think that minimizing pain (however defined) is not the right overall criterion - it should instead be to maximize effective participation (of which pain is certainly a factor but not the only one).   Granted, that quantity is even harder to define than pain.   But several people have observed that the meeting venue has a lot of effect on the ability to get work done.

Correct, and there was a WG mtgvenue that put out documents on this. Travel pain is one of the factors in deciding a venue. That's all I've been commenting on, but not b/c I think the other criteria should be ignored.


> Keith