[sigcomm] another approach

[Joe Touch]

Scott Shenker wrote:
> Here's a somewhat different discussion we could be having, motivated  
> by the following somewhat disjoint set of points:
> 
> 1)  I see the policy decision confronting SIGCOMM as roughly whether  
> or not SIGCOMM is willing to sponsor closed events and, if so,  
> whether or not those closed events can have any discretionary  
> attendance policies.  I think we are all in agreement that there  
> shouldn't be many closed events nor many discretionary attendees, so  
> the real question is whether SIGCOMM takes a stance that absolutely  
> precludes either.  I think we should reach closure (not consensus) on  
> this soon.

The stance has, FWIW, a range of "absolutes" that have been put forth
that distinguish different dimensions:

	- open vs. limited attendance
		- for limited: invitation vs. control-free attendance
	- one time vs. short-range vs. persistent limited attendance

Note that these affect HotNets because it is in the extrema of both
dimensions. The only other dimension we haven't talked about involves
more control over the papers themselves:

	- open vs. invitation-only presentations

That should be discussed as wekk.

> 2)  However, I think another, and more important, problem facing  
> SIGCOMM is the fact that our many of our conferences have so little  
> industry involvement.

It'd be useful to understand what kind of involvement we'd seek. It'd be
very useful, e.g., to have more interest in tech transfer of our ideas
to industry, or for industry to bring their problems to our community.

> 3)  Also, Hotnets was designed to with two goals in mind: (a)  
> encouraging "idea" papers and (b) fostering discussion during/after  
> the presentation.  These two goals need not be linked in other  
> conferences, and perhaps we should try to develop a new conference  
> with stresses (a) but not (b).

FDNA tried to do that, as do many of the other workshops at Sigcomm, FWIW.

> 4)  These two observations lead to the proposal that SIGCOMM start a  
> new multi-track conference that is open and invites 6-page papers  
> that are either "idea" papers (in the sense of Hotnets) or "reality"  
> papers (describing a current problem, or even just current practice  
> in a relevant area).  This could become the main alternative to  
> Sigcomm for the SIGCOMM community, providing a venue for the free  
> exchange of early ideas and where academia and industry could have  
> more fruitful interaction.

Sure - we had a survey about spinning FDNA in to a stand-alone meeting.
One trick is that the 'idea' papers from FDNA competed with the 'white
papers' sought for Sigcomm; we do need to keep that issue in mind. The
SIG tries not to sponsor or have 'in cooperation' meetings that compete
(addressing avoidance of competition is a prerequisite, as per the ACM
paperwork associated with this process).

> 5)  In addition to benefitting the SIGCOMM community, this would also  
> reduce the importance of the current controversy, in that if this new  
> conference was successful it would relegate Hotnets to the role of an  
> academic debating club (still available to those who enjoy that sort  
> of thing).

However, it would put the SIG in the position of picking one of the two
to sponsor or even endorse as 'in cooperation', as per above. That's the
catch here...

Joe

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 250 bytes
Desc: OpenPGP digital signature
Url : http://www.postel.org/pipermail/sigcomm/attachments/20051112/dc855806/signature.bin