From:
eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx
[mailto:eclipse.org-eclipsecon-program-committee-bounces@xxxxxxxxxxx] On
Behalf Of Bjorn Freeman-Benson
Sent: Tuesday, September 02, 2008 12:54 PM
To: Eclipsecon Program Committee list
Subject: Re: [eclipse.org-eclipsecon-program-committee] Yes/ No Polling
Analysis
Scott, et al,
Counting attendees is a good idea in theory, but not very workable in practice.
Consider the following difficulties:
- Multiple doors - we try to restrict inflow to one
door, but realistically people use other doors as well.
- People come and go - when do we count them? Do we
count only people who enter? What about people who leave to go to the
bathroom and then return? Double count?
- People who glance in, then decide not to stay - do
we count them?
- And the biggest problem: everyone streams into
the room in about three minutes. Getting an accurate count of 200+ people
coming through the door in three minutes is, um, probably not realistic.
- Bjorn
but I wonder if it would not be better to focus on getting accurate counts on
number of attendants for each session with the assumption that if someone
doesn't vote that they were neutral.
