[sigcomm] attendance policies for SIGCOMM-affiliated events

[Larry Peterson]

If the issue is a transparent invitation policy, then I don't think
anyone is arguing otherwise. However, this notion that the PC/chairs
are playing favorites and stacking the deck is both bewildering and
insulting.

In the interest of putting forward a tangible proposal that we can
discuss, I propose a transparent policy that roughly matches past
practice:

   o the PC and SC
   o one author per paper
   o as many students as we have scholarships for (preferring
     co-authors of accepted papers and then co-authors of
     submitted papers)
   o sponsor representatives (e.g., from NSF)

This gets us to roughly 80-90%. Then, at the Chair/PC's discretion:

   o as many second authors or authors of rejected papers as we can
     fit, perhaps with a slight bias for students/faculty at the hosting
     site.

Logistically, the invitation list probably needs to be finalized
by the chairs (as people accept or decline), but it seems reasonable
to have the PC put forward a list of "invite if room" people based
on the discussion during the PC meeting. (I should probably add at
this point that all HotNets have had co-chairs, meaning that there's
already some checks-and-balances built into the system.)

One final point. I know much of this discussion is about perception
as much as reality. Here's a small data-point of reality. In looking
at the registration list for 3 of the first 4 HotNets (all I have at
the moment), I count a total of 181 attendees and 145 unique  
individuals.

Larry

On Nov 8, 2005, at 9:01 PM, Joe Touch wrote:

>
>
> Vern Paxson wrote:
>
>>> ...
>>> This allows the chair to skew the content of the discussion.  
>>> E.g., if we
>>> want to have an ATM-fest, we might invite only ATM-friendly folk.  
>>> Or for
>>> ATM-bashing, only ATM haters.
>>>
>>
>> I must say, I was about this close to laughing out loud when I  
>> read this.
>> I have a hard time seeing this view as anything other than, well,  
>> paranoid.
>> Yes, what you describe is *possible* if the chair can invite  
>> attendees,
>> but it's by no means likely from my experience; and it's a  
>> presumption of
>> bad faith that I find disheartening.
>>
>
> It's an exaggeration to show the point. Proponents of invitation
> meetings already noted inviting people who they felt would be active
> participants (would they have a bias towards active 'pro'  
> participants,
> vs. active 'con' ones?), or limiting per-institution or per-group  
> (which
> presumes that all in a group or institution think the same way or have
> the same to offer a meeting).
>
> The key question is _why_ random isn't good enough, and why limit
> attendance at all?
>
>
>>> While I appreciate that we all would like to 'trust the chairs',  
>>> we do
>>> NOT trust the chairs to _invite_ all the papers (or the PC, for that
>>> matter)
>>>
>>
>> Regarding your parenthetical, in fact we do.  Case in point, this  
>> year's
>> SIGCOMM PC was selected by the chairs without input from the TAC.  We
>> requested that they solicit input; they declined; we still  
>> sponsored the
>> event.
>>
>
> The parenthetical was intended to be parsed the other way - neither  
> the
> chair nor the PC invite the papers.
>
> Joe
>
> _______________________________________________
> sigcomm mailing list
> sigcomm at postel.org
> http://www.postel.org/mailman/listinfo/sigcomm
>