Case Study in Large Theory Conferences: ICALP

Luca Aceto had posted a comment on my last post about STOC, describing his experiences running ICALP as a large theory conference. I thought it was fascinating, and requested that he repost a longer version on his blog. That post is here, and I strongly encourage anyone who's been involved in this discussion to go and read it... now....

I wanted to highlight a few snippets from it that I feel reinforce a number of points that I've been arguing.

ICALP is a three-track conference, and has been so since 2005, though typically only two tracks are running in parallel at each point in time. At ICALP 2008, in addition we had 12 satellite events, including the DYNAMO training school for doctoral students

Note that ICALP had an attendance range of about 500 - where we'd like STOC to be. It fits in with the pattern I was describing: more satellite events appears to correlate with larger attendance. 

As an aside, ICALP had Peter Winkler do a masterclass on math puzzles. Frankly, if we could just hire Peter Winkler and Persi Diaconis to do lectures at every STOC, our numbers would go into the stratosphere ;)

The workshops were held the day before ICALP or during the week-end following it. They were selected by the ICALP organizers amongst a fairly large number of proposals that we received in response to a call for workshops, based on their perceived scientific quality and on their potential interest to the ICALP community.

I've said this before, but I do think that if we go the route of having workshops/tutorials, the best way to do it is how other conferences do it: have a workshops chair solicit proposals, and decide from amongst them. The workshop organizers then take care of the work of getting speakers. It will ease the burden a lot.

I firmly believe that the task of organizing a conference like ICALP should be shared amongst several people. This certainly worked for us and helped us work more cheerfully, and overcome personal problems, mishaps and periods of crisis and panic that arose during the year before the conference took place

Very true. Again, most conferences that have lots of activities have a large organizing group, with proceedings chairs, arrangements chairs, workshop chairs, tutorial chairs, and so on. Apart from the fact that people's CVs get nicely bloated with service activities, more participation at the top level can actually help with overall attendance, as well as alleviating many of Michael's concerns (although technically he was more concerned with colocation, which is an idea I like, but does take more logistical coordination).