Re: Why not use on-line tool to track each working group's attendees?

[Joel Jaeggli]

On 8/6/10 9:48 AM, David Conrad wrote:
> Hi,
> 
> On Aug 6, 2010, at 4:36 AM, Ed Juskevicius wrote:
>> IEEE802 uses formal voting (during its meetings) to determine issues and to 
>> accept proposals (or motions) to add or delete text from documents that are 
>> being developed.
>> ...
>> The IETF WG process is very different from IEEE802, and that (for me) is 
>> reason enough to suggest that a system built for the IEEE802 may not be 
>> appropriate for IETF.  There are lots of other reasons too ...
> 
> OK, but the fact that our processes are different doesn't necessarily mean 
> moving to online tools is a bad idea, right?
> 
> Off the top of my head, I can see a couple of positives:
> - Online tools would allow folks who are participating remotely to indicate 
> 'presence' at a working group meeting.

Depending on the popularity of your working group and the extent to
which the meeting itself is timeshifted from the vantage point of the
bulk of the remote participants, as much as 10% of of the partficipants
might be remote, generally less, almost never more.

> - Simplify the administrative burden of collecting and keeping the blue sheets
> - Allow for easier analysis of working group attendance
> 
> And negatives:
> - Allow for easier analysis of working group attendance
> - Requires somebody to develop/maintain the tools
> - Requires everyone have some device (laptop, tablet, smartphone, etc) that 
> allows them to access the online tools (how does IEEE get around this?)
> - Might not meet the requirements (e.g., legal) the blue sheet currently fill

we have the rfid experiment as one example of an attendance
intrumentation exercise. I'm generally in favor of experimenting I think
it tells us as much about our current process as it does any change we
might be contemplating...

> I'm on the fence...
>
> Regards,
> -drc
> 
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www.ietf.org/mailman/listinfo/ietf
> 

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Yoav Nir]

Asia is big. Some parts of Asia (the middle east and the eastern parts of 
Russia) are closer to Europe than to China, Japan or Korea, at least as far as 
traveling goes.

But I think that only adds up to about 15-20 attendees, so it's still in the 
noise.

I also wonder if the data we have is about countries of citizenship, or where 
people actually work. A lot of people with Indian or Chinese  passports work in 
Silicon valley and are actually more comfortable with a meeting in north 
America than in Asia.

On Aug 8, 2010, at 3:04 AM, Michael StJohns wrote:


Hmm... folding Australia into Asia, Africa into Europe and S America into N 
America (for discussion purposes only) that's roughly

1/1/1.7 as a ratio. (Asia/Europe/NA).  Or 4/4/7.

It will be interesting to see what the other runs of the data show.

Mike



At 07:49 PM 8/7/2010, Donald Eastlake wrote:
Assuming the very simple model that attendance consists of a fixed number of 
constant attendees from each continent plus a "continentally local" variable 
number that only show up when the IETF meets on their continent and using the 
very limited data provided, using a rough least squares fit I get the following:

Constant Attendees
Africa 6
Asia 236
Europe 254
N.America 409
Australia 14
S.America 8

Continentally Local Attendees
Asia 333
Europe 173
N.America 232

Thanks,
Donald
==
 Donald E. Eastlake 3rd
 155 Beaver Street
 Milford, MA 01757 USA
 d3e...@gmail.com


On Fri, Aug 6, 2010 at 4:44 PM, Bob Hinden 
mailto:bob.hin...@gmail.com>> wrote:

During my IAOC chair plenary talk at IETF78 (slides are in the proceedings) I 
asked a question about continuing the current meeting policy (3 in North 
America, 2 in Europe, 1 in Asia in two year period (3-2-1) ) or changing to a 
1-1-1 policy based on current meeting attendance.  The talk included a graph of 
attendance by continent for IETF72-IETF78.  I was asked to provide this data to 
the community.

It is attached.  It includes the raw data and a new graph that shows attendance 
by percentage.  It appears to me that a 1-1-1 meeting policy is justified by 
current overall IETF meeting attendance.

Your comments are appreciated.

Bob







___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf


___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf



___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Bob Hinden]

Mike,

Just to be clear, the highest priority in venue selection is to find a venue 
where we can have a successful meeting.  We won't go anywhere were we don't 
think we can get the work done.  This discussion is where to have a meeting, 
but not at the expense of the work itself.

Bob

On Aug 7, 2010, at 4:15 PM, Michael StJohns wrote:

> Fred said this much more eloquently than I could.
> 
> On the IETF78 attendees list there's been a lot of discussion about where to 
> meet - with the primary consideration seeming to be "pretty and small".I 
> may be in the minority, but I'd really rather the IETF go places where the 
> ability to  "get work done" is the primary consideration.  
> 
> So going forward, I hope the considerations for location will give higher 
> weight to meeting the needs of the folks doing the work (my second list of 
> folk) and the folks who keep coming back (the first list) than to the single 
> meeting snap shots.  Its possible the demographics for my two lists are 
> similar to the raw demographics so my point may be moot - but why guess when 
> we have the data? 
> 
> Mike
> 
> 
> 
> At 12:34 AM 8/7/2010, Fred Baker wrote:
> 
>> On Aug 7, 2010, at 12:37 AM, Bob Hinden wrote:
>> 
>>> I do note that it seems clear that registration is related to where we 
>>> meet.  That show up pretty clearly the current data.  So judging where to 
>>> have future meetings based on past participation will tend to keep us where 
>>> we used to meet. Nomcom is, as you point out, 3 of 5 meetings.  WG chair 
>>> and authors might have a longer history.
>> 
>> I agree with the "openness" principle, but I disagree with this analysis. 
>> 
>> "3..5" is another way of saying "people that attend multiple times". As 
>> noted by others, first-time attendees (who by definition haven't attended 
>> anywhere else and therefore give us no guidance) and local-only attendees 
>> (which is unknowable but demonstrably a component) aren't very interesting. 
>> What is interesting is trying to serve people that participate. We went to 
>> Adelaide on the observation that we had IETF participation from there and a 
>> proposed host (which was also why Adelaide was chosen over, say, Sydney) at 
>> a time that we had never been to Australia. We went to Amsterdam, Stockholm, 
>> and so on on the observation that we had significant European participation 
>> and proposed hosts. We went to Japan when Japanese participation became 
>> important, and we're going to China in November largely in response to the 
>> fact of credible levels of Chinese participation. So observing participation 
>> doesn't limit us to where we have been, it extends us in the direction of 
>> those who par
 tic
>> ipate.
>> 
>> Looking at people who have attended multiple meetings, and using the nomcom 
>> rubric, make sense to me more than worrying about first-time and local-only 
>> attendees. I would take it on faith that we will have the latter wherever  
>> we go, and build on those that return.
>> ___
>> Ietf mailing list
>> Ietf@ietf.org
>> https://www.ietf.org/mailman/listinfo/ietf
> 
> 

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Tony Hansen]

1/1/2 comes closer to 1/1/1.7 than 1/1/1.

Tony Hansen
t...@att.com

On 8/7/2010 8:04 PM, Michael StJohns wrote:


Hmm... folding Australia into Asia, Africa into Europe and S America 
into N America (for discussion purposes only) that's roughly


1/1/1.7 as a ratio. (Asia/Europe/NA).  Or 4/4/7.

It will be interesting to see what the other runs of the data show.

Mike



At 07:49 PM 8/7/2010, Donald Eastlake wrote:
Assuming the very simple model that attendance consists of a fixed 
number of constant attendees from each continent plus a 
"continentally local" variable number that only show up when the IETF 
meets on their continent and using the very limited data provided, 
using a rough least squares fit I get the following:


Constant Attendees
Africa 6
Asia 236
Europe 254
N.America 409
Australia 14
S.America 8

Continentally Local Attendees
Asia 333
Europe 173
N.America 232

Thanks,
Donald
==
 Donald E. Eastlake 3rd
 155 Beaver Street
 Milford, MA 01757 USA
d3e...@gmail.com 


On Fri, Aug 6, 2010 at 4:44 PM, Bob Hinden > wrote:


During my IAOC chair plenary talk at IETF78 (slides are in the
proceedings) I asked a question about continuing the current
meeting policy (3 in North America, 2 in Europe, 1 in Asia in two
year period (3-2-1) ) or changing to a 1-1-1 policy based on
current meeting attendance.  The talk included a graph of
attendance by continent for IETF72-IETF78.  I was asked to
provide this data to the community.

It is attached.  It includes the raw data and a new graph that
shows attendance by percentage.  It appears to me that a 1-1-1
meeting policy is justified by current overall IETF meeting
attendance.

Your comments are appreciated.

Bob







___
Ietf mailing list
Ietf@ietf.org 
https://www.ietf.org/mailman/listinfo/ietf


___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf



___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf
   

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Fred Baker]

On Aug 7, 2010, at 4:15 PM, Michael StJohns wrote:
> I'd really rather the IETF go places where the ability to  "get work done" is 
> the primary consideration.  

To me, that's the only consideration apart from being open and spreading the 
travel pain among our participants.
___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Michael StJohns]

Hmm... folding Australia into Asia, Africa into Europe and S America into N 
America (for discussion purposes only) that's roughly

1/1/1.7 as a ratio. (Asia/Europe/NA).  Or 4/4/7.   

It will be interesting to see what the other runs of the data show.

Mike



At 07:49 PM 8/7/2010, Donald Eastlake wrote:
>Assuming the very simple model that attendance consists of a fixed number of 
>constant attendees from each continent plus a "continentally local" variable 
>number that only show up when the IETF meets on their continent and using the 
>very limited data provided, using a rough least squares fit I get the 
>following:
>
>Constant Attendees
>Africa 6 
>Asia 236 
>Europe 254 
>N.America 409 
>Australia 14 
>S.America 8 
>
>Continentally Local Attendees
>Asia 333 
>Europe 173 
>N.America 232 
>
>Thanks,
>Donald
>==
> Donald E. Eastlake 3rd
> 155 Beaver Street
> Milford, MA 01757 USA
> d3e...@gmail.com
>
>
>On Fri, Aug 6, 2010 at 4:44 PM, Bob Hinden 
><bob.hin...@gmail.com> wrote:
>During my IAOC chair plenary talk at IETF78 (slides are in the proceedings) I 
>asked a question about continuing the current meeting policy (3 in North 
>America, 2 in Europe, 1 in Asia in two year period (3-2-1) ) or changing to a 
>1-1-1 policy based on current meeting attendance.  The talk included a graph 
>of attendance by continent for IETF72-IETF78.  I was asked to provide this 
>data to the community.
>
>It is attached.  It includes the raw data and a new graph that shows 
>attendance by percentage.  It appears to me that a 1-1-1 meeting policy is 
>justified by current overall IETF meeting attendance.
>
>Your comments are appreciated.
>
>Bob
>
>
>
>
>
>
>
>___
>Ietf mailing list
>Ietf@ietf.org
>https://www.ietf.org/mailman/listinfo/ietf
>
>
>___
>Ietf mailing list
>Ietf@ietf.org
>https://www.ietf.org/mailman/listinfo/ietf

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Lixia Zhang]

On Aug 7, 2010, at 4:15 PM, Michael StJohns wrote:

> Fred said this much more eloquently than I could.
> 
> On the IETF78 attendees list there's been a lot of discussion about where to 
> meet - with the primary consideration seeming to be "pretty and small".I 
> may be in the minority, but I'd really rather the IETF go places where the 
> ability to  "get work done" is the primary consideration.  

I'd like to second Mike here.
admittedly I'm a boring type, my preference is a more or less fixed set of 
meeting locations (Minneapolis is just fine), one knows exactly what to expect, 
easy to plan. Perhaps easier on IAOC as well.

However one of the few fundamental truths about Internet is its diversity, that 
applies to the community as well, hence those who desire pretty and small. 
I expect we end up in some kind of engineering decisions, trying best to 
accommodating different interest with well informed tradeoffs.

Lixia

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Donald Eastlake]

Assuming the very simple model that attendance consists of a fixed number of
constant attendees from each continent plus a "continentally local" variable
number that only show up when the IETF meets on their continent and using
the very limited data provided, using a rough least squares fit I get the
following:

Constant Attendees
 Africa 6  Asia 236  Europe 254  N.America 409  Australia 14  S.America 8
Continentally Local Attendees
 Asia 333  Europe 173  N.America 232
Thanks,
Donald
==
 Donald E. Eastlake 3rd
 155 Beaver Street
 Milford, MA 01757 USA
 d3e...@gmail.com


On Fri, Aug 6, 2010 at 4:44 PM, Bob Hinden  wrote:

> During my IAOC chair plenary talk at IETF78 (slides are in the proceedings)
> I asked a question about continuing the current meeting policy (3 in North
> America, 2 in Europe, 1 in Asia in two year period (3-2-1) ) or changing to
> a 1-1-1 policy based on current meeting attendance.  The talk included a
> graph of attendance by continent for IETF72-IETF78.  I was asked to provide
> this data to the community.
>
> It is attached.  It includes the raw data and a new graph that shows
> attendance by percentage.  It appears to me that a 1-1-1 meeting policy is
> justified by current overall IETF meeting attendance.
>
> Your comments are appreciated.
>
> Bob
>
>
>
>
>
>
>
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www.ietf.org/mailman/listinfo/ietf
>
>
___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Michael StJohns]

At 09:05 AM 8/7/2010, Marshall Eubanks wrote:
>Dear Noel;
>
>On Aug 6, 2010, at 9:26 PM, Noel Chiappa wrote:
>
>>> From: Bob Hinden 
>> 
>>> I do note that it seems clear that registration is related to where
>>> we meet. That show up pretty clearly the current data. So judging
>>> where to have future meetings based on past participation will tend
>>> to keep us where we used to meet.
>>> ...
>>> I think an important part of the meeting rotation is to equalize the
>>> travel cost/pain for most attendees. 
>> 
>> The last makes some sense, but I wonder about the 'local attendees'
>> affect. Clearly you will always get a goodly number of people from the
>> location where the meeting is, but how far does the 'continental' effect
>> reach in that breakdown?
>
>We do have some data on this point - the day pass experiment (DPE) has shown 
>pretty conclusively IMO that the IETF
>does not get a lot of truly local ad-hoc participants. Most day pass attendees 
>appear to be regular attendees who could only make it to that particular IETF 
>for one day for whatever reason, not local people who just wanted to sample an 
>IETF meeting. 
>
>It has long been known that IETF meetings have a local attendance effect. I 
>thought, before the DPE, that this indicated a potentially large number of 
>observers, presumably interested, but not interested enough to travel long 
>distances due to the cost and time required for longer trips. This, to me, 
>suggested that day-passes, at a reduced rate, would bring out a lot of new 
>people (as the time and financial burden would be even less). This did not 
>happen, on any of the 3 continents where the DPE has been run. 
>
>So, I now assume that the "local attendees" are people who are seriously 
>interested and involved in the IETF, able to travel in-region or in-country 
>but unable to get approval, funds or time for week-long international travel. 


I hate to be blunt, but this reasoning appears seriously flawed.  

None of the last 3 sites have been what I would consider "convenient" for drop 
ins.  Had these been in say San Jose, or DC, or London or Tokyo I might agree, 
but not for the sites actually chosen.  When you get to a certain distance from 
your home, a day pass makes no economic nor time commitment sense - better to 
attend the most of the week (or at least a couple of days), do your working 
group meetings, attend the plenaries and have dinner with your colleagues (or 
the folks you just met in the WG meeting) than try and fly in on one day, do a 
1-2 hour WG and then decamp.  I've done that - but its not the best use of my 
time.

I mean really - is Maastricht such a hot bed of Internet design activity that 
you would expect 20-30 locals to show up?  And being 2 hours from Brussels and 
3 from Amsterdam, I wouldn't expect to pick up any day pass folk from there.  I 
forget what your numbers showed, but weren't most of the day pass folk previous 
attendees?

I wouldn't generally expect first-timers to use a day pass. Mostly, they don't 
understand what an IETF meeting is like (and generally don't believe us when we 
tell them).  Many of the first timers I've met over the years are surprised 
there aren't more instructional sessions - they're thinking symposium, not 
working meeting - and they're here for the whole learning experience.  I would 
expect day passes to be more attractive to a repeat attendee involved in a 
specific working group who has other time commitments dragging him away from 
the site (work, family, etc).

Mike


___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Michael StJohns]

Fred said this much more eloquently than I could.

On the IETF78 attendees list there's been a lot of discussion about where to 
meet - with the primary consideration seeming to be "pretty and small".I 
may be in the minority, but I'd really rather the IETF go places where the 
ability to  "get work done" is the primary consideration.  

So going forward, I hope the considerations for location will give higher 
weight to meeting the needs of the folks doing the work (my second list of 
folk) and the folks who keep coming back (the first list) than to the single 
meeting snap shots.  Its possible the demographics for my two lists are similar 
to the raw demographics so my point may be moot - but why guess when we have 
the data? 

Mike



At 12:34 AM 8/7/2010, Fred Baker wrote:

>On Aug 7, 2010, at 12:37 AM, Bob Hinden wrote:
>
>> I do note that it seems clear that registration is related to where we meet. 
>>  That show up pretty clearly the current data.  So judging where to have 
>> future meetings based on past participation will tend to keep us where we 
>> used to meet. Nomcom is, as you point out, 3 of 5 meetings.  WG chair and 
>> authors might have a longer history.
>
>I agree with the "openness" principle, but I disagree with this analysis. 
>
>"3..5" is another way of saying "people that attend multiple times". As noted 
>by others, first-time attendees (who by definition haven't attended anywhere 
>else and therefore give us no guidance) and local-only attendees (which is 
>unknowable but demonstrably a component) aren't very interesting. What is 
>interesting is trying to serve people that participate. We went to Adelaide on 
>the observation that we had IETF participation from there and a proposed host 
>(which was also why Adelaide was chosen over, say, Sydney) at a time that we 
>had never been to Australia. We went to Amsterdam, Stockholm, and so on on the 
>observation that we had significant European participation and proposed hosts. 
>We went to Japan when Japanese participation became important, and we're going 
>to China in November largely in response to the fact of credible levels of 
>Chinese participation. So observing participation doesn't limit us to where we 
>have been, it extends us in the direction of those who parti
 c
> ipate.
>
>Looking at people who have attended multiple meetings, and using the nomcom 
>rubric, make sense to me more than worrying about first-time and local-only 
>attendees. I would take it on faith that we will have the latter wherever  we 
>go, and build on those that return.
>___
>Ietf mailing list
>Ietf@ietf.org
>https://www.ietf.org/mailman/listinfo/ietf


___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Michael StJohns]

At 06:50 PM 8/6/2010, DOLLY, MARTIN C (ATTLABS) wrote:
>Though interesting, what is the intent of the use of this data
>Martin
>Martin C. Dolly

Seems pretty obvious - same intent as the original data.  Determining if a 
1-1-1 model is appropriate. 



___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: DPE and Nomcom [was Re: IETF Attendance by continent]

[Marshall Eubanks]

On Aug 7, 2010, at 5:35 PM, Brian E Carpenter wrote:


On 2010-08-08 03:11, Doug Ewell wrote:

Marshall Eubanks  wrote:

We do have some data on this point - the day pass experiment (DPE)  
has

shown pretty conclusively IMO that the IETF does not get a lot of
truly local ad-hoc participants. Most day pass attendees appear to  
be

regular attendees who could only make it to that particular IETF for
one day for whatever reason, not local people who just wanted to
sample an IETF meeting.

It has long been known that IETF meetings have a local attendance
effect. I thought, before the DPE, that this indicated a potentially
large number of observers, presumably interested, but not interested
enough to travel long distances due to the cost and time required  
for

longer trips. This, to me, suggested that day-passes, at a reduced
rate, would bring out a lot of new people (as the time and financial
burden would be even less). This did not happen, on any of the 3
continents where the DPE has been run.


At least on the surface, this does make it appear that the decision  
to
exclude day-pass attendees from NomCom, on the basis that such  
attendees

would not have the requisite experience, is driven by financial
considerations after all.


Not at all. Firstly, it's only now that there are even marginally  
enough

data to analyse the impact of day passes - previously, all we had
was guesswork. Secondly, the argument is quite clear - people who
parachute into an IETF week for one day, for whatever reason, cannot
possibly gain the experience of a full week's participation (for
example, witness the breadth and depth of an AD's work, or properly
understand the interaction of WG meetings, BOFs, Bar BOFs and
corridor discussions). That undermines the whole purpose of measuring
meeting attendance as a Nomcom qualification.


I agree. Note also that the number of day passes per meeting has been  
so small that the financial effect of

the day pass experiment is literally in the noise at present.

Regards
Marshall



  Brian
___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf



___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

DPE and Nomcom [was Re: IETF Attendance by continent]

[Brian E Carpenter]

On 2010-08-08 03:11, Doug Ewell wrote:
> Marshall Eubanks  wrote:
> 
>> We do have some data on this point - the day pass experiment (DPE) has
>> shown pretty conclusively IMO that the IETF does not get a lot of
>> truly local ad-hoc participants. Most day pass attendees appear to be
>> regular attendees who could only make it to that particular IETF for
>> one day for whatever reason, not local people who just wanted to
>> sample an IETF meeting.
>>
>> It has long been known that IETF meetings have a local attendance
>> effect. I thought, before the DPE, that this indicated a potentially
>> large number of observers, presumably interested, but not interested
>> enough to travel long distances due to the cost and time required for
>> longer trips. This, to me, suggested that day-passes, at a reduced
>> rate, would bring out a lot of new people (as the time and financial
>> burden would be even less). This did not happen, on any of the 3
>> continents where the DPE has been run.
> 
> At least on the surface, this does make it appear that the decision to
> exclude day-pass attendees from NomCom, on the basis that such attendees
> would not have the requisite experience, is driven by financial
> considerations after all.

Not at all. Firstly, it's only now that there are even marginally enough
data to analyse the impact of day passes - previously, all we had
was guesswork. Secondly, the argument is quite clear - people who
parachute into an IETF week for one day, for whatever reason, cannot
possibly gain the experience of a full week's participation (for
example, witness the breadth and depth of an AD's work, or properly
understand the interaction of WG meetings, BOFs, Bar BOFs and
corridor discussions). That undermines the whole purpose of measuring
meeting attendance as a Nomcom qualification.

   Brian
___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Doug Ewell]

Marshall Eubanks  wrote:

We do have some data on this point - the day pass experiment (DPE) has 
shown pretty conclusively IMO that the IETF does not get a lot of 
truly local ad-hoc participants. Most day pass attendees appear to be 
regular attendees who could only make it to that particular IETF for 
one day for whatever reason, not local people who just wanted to 
sample an IETF meeting.


It has long been known that IETF meetings have a local attendance 
effect. I thought, before the DPE, that this indicated a potentially 
large number of observers, presumably interested, but not interested 
enough to travel long distances due to the cost and time required for 
longer trips. This, to me, suggested that day-passes, at a reduced 
rate, would bring out a lot of new people (as the time and financial 
burden would be even less). This did not happen, on any of the 3 
continents where the DPE has been run.


At least on the surface, this does make it appear that the decision to 
exclude day-pass attendees from NomCom, on the basis that such attendees 
would not have the requisite experience, is driven by financial 
considerations after all.


--
Doug Ewell | Thornton, Colorado, USA | http://www.ewellic.org
RFC 5645, 4645, UTN #14 | ietf-languages @ is dot gd slash 2kf0s ­

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf

Re: IETF Attendance by continent

[Marshall Eubanks]

Dear Noel;

On Aug 6, 2010, at 9:26 PM, Noel Chiappa wrote:

>> From: Bob Hinden 
> 
>> I do note that it seems clear that registration is related to where
>> we meet. That show up pretty clearly the current data. So judging
>> where to have future meetings based on past participation will tend
>> to keep us where we used to meet.
>> ...
>> I think an important part of the meeting rotation is to equalize the
>> travel cost/pain for most attendees. 
> 
> The last makes some sense, but I wonder about the 'local attendees'
> affect. Clearly you will always get a goodly number of people from the
> location where the meeting is, but how far does the 'continental' effect
> reach in that breakdown?

We do have some data on this point - the day pass experiment (DPE) has shown 
pretty conclusively IMO that the IETF
does not get a lot of truly local ad-hoc participants. Most day pass attendees 
appear to be regular attendees who could only make it to that particular IETF 
for one day for whatever reason, not local people who just wanted to sample an 
IETF meeting. 

It has long been known that IETF meetings have a local attendance effect. I 
thought, before the DPE, that this indicated a potentially large number of 
observers, presumably interested, but not interested enough to travel long 
distances due to the cost and time required for longer trips. This, to me, 
suggested that day-passes, at a reduced rate, would bring out a lot of new 
people (as the time and financial burden would be even less). This did not 
happen, on any of the 3 continents where the DPE has been run. 

So, I now assume that the "local attendees" are people who are seriously 
interested and involved in the IETF, able to travel in-region or in-country but 
unable to get approval, funds or time for week-long international travel. 

Regards
Marshall

> 
> E.g. for a North American meeting on the West Coast, how many North
> Americans are coming from the East Coast? For the Far East, how many are
> coming from other countries in the Far East?
> 
> To put it another way, do we have a substantial pool of people who will
> travel a long ways for a meeting in their home continent, but not to other
> continents? I think that's the group (and its breakdown by continent) we
> really need to look for/at. _If_ that's heavily skewed to one continent,
> we might want to bias the meeting schedule in their favour.
> 
> (The people who will always come, no matter where it is, aren't as much of
> an issue, although all else being equal, as you point out it would be good
> to equalize their pain. As for the local-only attendees, well, no matter
> where you go, you only get them from close by, and we can't visit all
> plausible cities on any sort of schedule, so I'm not sure how much weight
> we want to put on letting one-time attendees get a taste of the IETF.)
> 
>   Noel
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www.ietf.org/mailman/listinfo/ietf
> 

___
Ietf mailing list
Ietf@ietf.org
https://www.ietf.org/mailman/listinfo/ietf