I don't think we have a specific set of requirements from the
community. Dave Crocker and some friends started putting some notes
together more than a decade ago, and last I spoke with him about it he
had dropped it because he had been unable to form a consensus. Maybe
that needs to be renewed. What we do understand we know as much as
anything because we ourselves are IETF participants. As Bob put it,
the first requirement is that the meeting be effective, and the direct
implications of that include that it be reasonably accessible at a
rational cost, that it have a meeting facility with an appropriate set
of rooms, that it have enough hotels and restaurants and that they be
rationally located with respect to the meeting facility, and so on.
After that, we get into a lot of minutia of the costs of this and that
and the desires of the sponsors. There are also some purely business
issues that come up - sometimes we can horse-trade prices when
comparing locations.
Taking one example that you mention in your note, I don't believe that
the IAOC understands the community requirement to be "direct
international flight" - we don't bend a rule there, we understand the
rule to be something different. If we did think the rule was "direct
international flight" and interpreted that literally - "for any given
meeting, a participant from Australia, New Zealand, Korea, China,
Japan, any European country, Canada, or the US needs to be able to get
to the exact location using at most one international flight and take
no additional leg afterwards" - I think the set of places we could
meet would be very small, probably the null set, simply because of the
number of places we come from and their various geographies. What we
do understand the requirement to be is that the number, cost, and
convenience of connections, along with other costs and issues, be
optimized for the average participant, and yes we try to do that to
the extent that we can.
In consideration of that, look at the question we asked about
Vancouver vs Quebec City. Ray will give the results Monday when the
poll closes, but at this instant we are running about 2:1 in favor of
Quebec City. For our Asian colleagues, that means flying to a major US
or Canadian airport like EWR, ORD, or YYQ, and taking another flight
to YQB. For our European colleagues, there are a few direct flights
from Europe, but for the most part they will similarly go to New
Jersey, Chicago, Montreal, or Toronto, and then on to YQB; from ARN to
YQB int.net finds 44 paths but only five that have a single stop, and
those are all via ORD or EWR. In the US, folks will fly to one of a
few major hubs and fly on to YQB. For me personally, itn.net tells me
that I will fly from SBA to a major hub like LAX or SFO, from there to
a hub near the US/Canadian border like ORD, DTW, YYZ, or EWR, and then
proceed to Quebec City - three legs, and I will probably do it the day
before rather than trying to do it in one day. I would expect that our
ANZ colleagues will follow me - a long hop to LAX or SFO, another to
the US border with Canada, and a final hop. Vancouver is a major
international hub, but Quebec City is not; like Minneapolis, Dublin,
Prague, San Diego, Dallas, Vienna, Stockholm, Hiroshima, Atlanta,
Orlando, and Maastricht, it is a place one gets to one hop beyond a
major hub. And what we are being told is - "Vancouver is nice, but
let's go to Quebec City".
On May 28, 2009, at 12:39 PM, John C Klensin wrote:
Bob,
I'd like to express my thanks for this information. I believe
that, if the community had more information of this sort,
updated as needed, we would have fewer firestorms when meeting
sites are announced.
When you have time, I (and I believe others) would like to
understand better how you evaluate "reasonable costs for the
IETF and attendees". I think it is general knowledge that it
is possible to trade IETF costs off against participant costs
(e.g., making things cheaper for the IETF and more expensive for
participants). It would be good to know how the participant
costs are estimated and how the two are balanced.
Similarly, there is presumably a point at which sponsor
preferences have to give way to other considerations. Could you
give us a sense as to when that occurs? Put differently, at
what point might you decide that dealing with a particular
potential sponsor's preferences are sufficiently inflexible
and/or onerous that you lose interest in sponsorship from that
organization? It appears to me, for example, that the IAOC has,
more than once, dropped both the "same facility" and "direct
international flights" requirements, and pushed the boundaries
of the "participant costs" one in order to get a sponsor and
accommodate their needs. That isn't necessarily the wrong
decision, but I believe the community should be better
informed... and that we would have much less
ruckus if it were.
regards and thanks again,
john
--On Wednesday, May 27, 2009 22:15 -0700 Bob Hinden
<bob.hinden@xxxxxxxxx> wrote:
Given the recent discusion on travel to IETF78, I wanted to
summarize how the IAOC makes decisions on IETF meeting
locations.
We first set the dates and general geographic location about
three to four years in advance. For example, the future
meeting calendar at
http://www.ietf.org/meetings/0mtg-sites.txt goes out through
the end of 2013. We are maintaining on average a rotation
between Asia, North America, and Europe with the goal of no
more than two of three meetings in North America. We do this
because we believe the community wants the dates set well in
advance and to have a reasonable balance between NA, Europe,
and Asia.
The IAOC's most important criteria for meeting locations is
where can we have a successful IETF meeting. That is, hold a
meeting where the IETF can get its work done. The next items
are 1) having the meeting and hotel rooms in the same building
or a conference center next to the
hotels, 2) direct international flights, 3) reasonable costs
for the IETF and attendees, and 4) the ability to deploy an
IETF style network in the venue. This list isn't ordered and
we don't always achieve all of this for every meeting, but
these are the goals. There are always tradeoffs.
If we have a host for a specific meeting and the host has a
location preference we try to honor their preference. Having a
host for a meeting significantly reduces the cost to have an
IETF meeting. This is especially important for meetings
outside of North America where the meetings rooms are not
included with a block of hotel rooms. Due to the current Visa
situation in the US we are planning many meetings outside of
the US.
In the case of hosted meetings we might relax some of our
criteria like direct flights or hotel and meeting rooms in the
same building, but we will never pick a venue where we don't
think we can have a successful IETF meeting. For non-hosted
...
_______________________________________________
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf