[ga] Selecting Comments to Read Aloud

["Ben Edelman" <edelman@law.harvard.edu>]

I thought at some length about Kendall's 2/13 suggestion (in short, that
remote participants "quickly vote on 2-3 questions which would then be asked
out-loud") and discussed it with the rest of the Berkman Center team that
would actually be responsible for implementing such a proposal.

This seems to us like a good idea, and we're hopeful that we can pull off
something along the lines of what Kendall proposes.  To be precise, we feel
reasonably confident about implementing the scripts to support this
suggestion, devising appropriate fraud control systems to guard against the
most obvious opportunities for deception, and actually running the system in
a way that is successful, accomplishing the intended goals without
distracting our staff from other important responsibilities.

Accordingly, I expect that this system will be used in the 3/11 DNSO GA
session, and likely in the NC session to follow (with the permission of the
NC chair).  Its use in the 3/12 Public Forum will surely be contingent on a
number of factors -- in particular, a positive 3/11 evaluation by our team
(for reliability as well as for effect on the workload of our staff) and by
remote participants, and approval of Mr. Cerf, in his capacity as Chairman
of the ICANN Board and thus chair of the 3/12 Public Forum.

(For whatever it's worth, my initial estimates for the project anticipate
that endorsements will triple the number of database write operations, and
may increase read operations by a factor of ten to twenty, depending on
behavioral response [i.e. folks repeatedly pressing Refresh to see updated
votes] which are both potentially of some concern.  Also serious is the fact
that I'll be writing quite a bit of fresh code for this purpose --
rigorously tested, to the extent possible, but even so unused in the real
world -- and in fact anticipate doubling the number of lines of code
supporting the remote participation process in order to implement the
peculiar mechanisms that seem to be called for here.  So while I expect
success in 3/11, there's of course no guarantee, and if things should go
badly on 3/11, we'll definitely return to the "backup plan" [of the current
system] on 3/12.)


Further comments and suggestions about the remote participation system
remain in order at this time.  We're still working on final tweaks, and a
sizable portion of improvements result directly from suggestions made by you
folks, as remote participants in prior meetings.


Ben Edelman
Berkman Center

--
This message was passed to you via the ga-full@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga-full" in the body of the message).
Archives at http://www.dnso.org/archives.html