21 years in the sausage business

Hey, kids! It's time for the business meeting for the 21st Annual ACM Symposium on Computational Geometry! Let's listen to what those wacky geometers are doing in Pisa this year.

There's no beer! There's no wine! What's wrong with these Italian teetotalers?

Pankaj thanks the committees, everyone claps.

Beppe Liotta reports on local organization: 120 participants (102 pre-registered) from 18 countries, plus Australia. Specifically: 22 US, 20 Germany, 13 Italy, 11 France, 8 S. Korea.... The European-standard fees were way too high, but it's easy to see why. Income: €28,000 from fees, €3900 from Raindrop, €3000 from IIT. Expenses: €19,620 (!!!) for organization expenses (hiring the company that organized the conference, renting room, invited speakers, PC dinner) plus €3500 for proceedings, plus 5% Visa fees, 20% VAT, 18% ACM service fees. We barely broke even. Beppe thanks the local volunteers. Everyone claps. Then lots of people start complaining about hiring an expensive organizing company, but apparently, “that's just how things are done here in Italy.” Eventually, Pankaj cuts off the increasingly strident complaints so we could eat later.

Günter Rote gives the program committee report. First, he points out that this is the 20th anniversary meetings, and that the name of the conference does NOT actually contains the word "ACM". Günter threw up some invisible (but presumably very important) statistics, which were still invisble after he increased the font size from "standard" to "larger", despite screams of "largEST!" from your intrepid reporter. The overall acceptance rate was 29%. Most papers were about 2d problems. Most papers were about points; higher-degree algebraic curves were least successful. There were many different problem areas, none significantly more successful than others. Papers with long history are more successful than preliminary work, but new solutions are better than old ones. Papers with low-order polynomial running times were popular; papers with sub-constant running times had a perfect success rate (0/0). Papers whose authors promised never to implement their algorithm had a perfect acceptance rate (0/0). The most successful application area was cartography and GIS. Papers with student co-authors were less successful; papers with industry co-authors were more successful. All this data is available somewhere on the web, but Günter diidn't give us the URL, so you can't see them, so there. Günter also showed some nice cartograms, generated by Bettina Speckmann, but you can't see those, either. The shortest submission was one page long (with six lines of text). Two papers were withdrawn, yay.

Where is my beer?

Joe Mitchell asks everyone on the committee to stand up. Everybody claps. Joe gives another breakdown of submissions: 2 withdrawn (yay) + 97 theory + 16 theory with experiments + 20 "applied" + 8 "left field" = 143. Joe's personal opinion was that 90 submissions were "acceptable" in the absense of schedule constraints (although the PC might only agree on 65); only 41 papers were actually accepted. Each PC member was assigned about 45 submitted papers. 32% of submitting authors would prefer to submit papers in final proceedings format; the others either preferred the usual single-column format or didn't care. A long drawn-out lack-of-alcohol-induced discusssion ensues. Blah blah blah. Where the hell is my beer? Oh, good, Pankaj is cutting off the discussion. Everybody claps.

Lutz Kettner talks about the video/multimedia submissions. "Okay let's keep this brief." Yay! There were 13 submissions in several formats (all electronic), 11 of which were accepted, tying the previous record in 1999 when some idiot with long hair was in charge of the committee. There were no wardrobe malfunctions, and none of the videos were actually subliminal advertisements for beer. Mmmm, beer. Cold, bubbly beer. There is an online proceedings, soon to appear at videos.compgeom.org. The multimedia program was produced without the help of Pauwel Kwak. The committee preferred PAL over NTSC, because (unlike people who organize beer-free business meetings) pixels are not square. Lutz suggest limiting the accepted file/video submision formats in future years: pilsner, hefeweizen, stout, and lambic only—certainly none of that mamby-pamby canoe-sex American light lager nonsense. I am really thirsty. Oh, hey, everyone's clapping again.

Nina Amenta and Otfried Cheong are the PC chairs next year. The other members of the committee are Dominique Attali, Peter Braß, Timothy Chan, Ken Clarkson, Sylvain Lazard, Joe O'Rourke, Rom Pinchasi, Sylvani Pion, Emil Praun, Raimund Seidel, Alper Üngör, Antoine Vigneron, Yusu Wang, and Sue Whitesides. Nina and Otfried are planning a physical PC meeting (presumably with beer) for the first time in many many years. The PC is considering allowing authors to rebut negative reviews, as at SIGGRAPH; we'll see.

Joe Mitchell puts on his Alon Efrat costume to talk about Sedona, Arizona, the center of the Harmonic Convergence and the site of next year's conference, to be held June 5–7, 2006. Because it's a dry heat™, Sedona is one of the heaviest consumers of beer in the entire United States. While enjoying your hoppy beverage, you can also enjoy some incredibly beautiful scenery. Flights will be a bit difficult, since Sedona doesn't have amajor airport. The Sedona Hilton is a nice hotel, with rooms costing $147 a night (eek). Sedona is a popular tourist destination, so if you want to book a cheaper hotel room (so you can afford more beer), you should book early.

Jack Snoeyink says to submit videos and then sits down. Everyone mentally applauds Jack's brevity and fashion sense.

There were four bids for SOCG 2007:

• 2007 is an FCRC year, but if we're sane (and we were), we won't care (and we don't). According to Ken Clarkson, FCRC will be in San Diego (Gateway to San Diego County), June 9–19, in the same incredibly boring hotel as in 2003. Completely refurbished! With wireless everywhere! Near the tram! Far away from everything except the mall! Expensive! But at least they'll have beer, and Tijuana is only an hour away by tram.

• Otfried Cheong (né Schwarzkopf) bids for his adoptive homeland, Korea. It's difficult (plane+plane+bus or plane+train+train) and expensive to get there. But there are lots of local students, cheap hotels, cheap food, and government funding for conferences. Korea is the only country in the world with a state-funded Voronoi Diagram Research Center. In 2007, JCDCG will be in mid-June to celebrate Jin Akiyama's (n+1)th birthday. Korea is not in the southern hemisphere. Otfried wants to have the conference in Gyongju (Kyŏngju), the historical capital in the 7th to 10th century (like Kyoto). Lots of temples; lots of stone Buddhas. Namsan Mountain looks really cool.

• Joe Mitchell puts on his David Mount costume and bids for Alexandria, VA, near Washington, DC. The weather usually sucks in DC, "David" says, "but maybe we'll luck out." Joe doesn't take credit for David's jokes, which went by too fast for me to copy. The cheaper and less nice option (since Alexandria is expensive) is the College Park campus, but then it'll have to be in the last week of May to get rooms on campus. Maybe we'll cruise on the Potomac, drinking beer and watching the government sell off the monuments on eBay to pay off the national debt.

• Finally, Jack Snoeyink (né Snoeyink) bids for his adoptive homeland, Chapel Hill, North Carolina. He wears a snazzy jacket and carries a snazzy tablet PC, which would be really cool if only it didn't run Windows, because then we wouldn't have to wait for the PowerPoint presentation to come up. (The buses to the banquet have been delayed 15 more minutes to let us finish the meeting. Damn it.) Old university town, cheap eats, lots of beer. LOTS of beer. Join me, won't you, while I fondly remember what little I still can of Top of the Hill and the Carolina Brewery. Mmm. Beer. Lovely, bubbly, thirst-quenching, micro-brewed beer. Oh yeah, and it'll be very very cheap. The banquet will be an outside barbecue, like at Berkeley (woo) and Vancouver (yay). Organizers will be student volunteers; student registrations fees withh be $100 or two bottle caps. Conference attendees will be given free tickets to the annual UNC/Duke basketball game, which will be delayed especailly for SOCG if we decide to attend.

• Pankaj calls for the first vote: for FCRC versus against FCRC. No wait, Micha wants to vote differently. No wait, Vladlen asks how we can vote. Okay, fine, whatever. The vote was 8 for FCRC, 57 for Korea, 3 for DC, and 38 for North Carolina. (Ken Clarkson mutters, "Have you been to Norht Carolina? Holy crap!") Runoff? No, Korea has a strict majority! Everyone claps! It's the fastest vote in history! Maybe sobriety isn't so bad after all. On the other hand, now we won't get to watch Duke beat UNC, unless we can get both teams to come to Gyongju.
  

Pankaj mentions two leftover issues in the last five minutes of the meeting. Shoudl we accept more papers? Should we change the submission format? Ah, whatever, let the PC and steering committee work it out. No really, everybody, shut up. Oh, good, everyon'e clapping again. Time to catch the bus.

Comments

[Anonymous]

Many have said it before, and I'll say it again. It hurts the field to claim that the are only 41 worthy papers in Computational Geometry.

How can we ask the NSF for extra funding claiming that Comp. Geom. is a hotbed of activity when our main conference has only 41 papers? Contrast this with, say SIGGRAPH. Who can argue that there is interest in that field on the face of their wildly successful conference?

I know of at least one case where the paper was rejected in SoCG and accepted in STOC (!!).

[Andrea]

Great summary! (Pros: can choose to drink beer while reading summary. Cons: Not in Italy.)

[John]

I love statistics. Here are some to ponder.

Statistics for the past few SOCG's:

Attend ACM/Student Location
2005 120 E 320 / 176 Pisa (no banquet, add E 60, no student discount)
2004 180 $ 260 / 98 Brooklyn
2003 124 $ 330 / 155 FCRC (no banquet)
2002 143 E 200 / 100 Barcelona
2001 153 $ 285 / 125 Tufts
2000 129 Hong Kong

It seems the more you pay, the smaller the attendance, and
the banquet is removed.

This year there were E20,000 (E167 per person) in expenses excluding
overhead, VAT, credit card fees, proceedings and banquet. Not included
were meals, banquet, food at the coffee break, or cake or beer at the
business meeting.

For comparison, let's see what would have been charged last year without
the banquet.

Total (per person)

$4547 ($25) Coffee Breaks (6, with food)
$4160 ($23) Welcome reception
$2340 ($13) Invited speakers (1 local, one not, same as this year)
$1157 ($ 6) PC Dinner (Nobu)
$ 638 ($ 4) 4 color pen, paper, folder, name tag
$ 207 ($ 1) Beer, ice, tubs for beer (Yes, there was beer!)
$ 105 ($ 0.58) Flour, sugar, etc for Herve to make a cake for the business meetings
$ 8 ($ 0.04) Ice (for beer)
$ 4 ($ 0.02) Bottle openers (for beer)
$ 0 ($ 0) Room rental, general office supplies, copying (e.g. the program), secretarial support,
students to work the registration desk, projectors, amplification,
custodial support, internet.

Total: $13,162 ($73 per person)

Of these expenses, $9,000 was covered by Raindrop and CATT,
our sponsors. There was about the same level of sponsorship
this year.

Total (after sponsors): $ 4,162 ($23 per person)

Thus, adding ACM's 18% tax, the proceedings ($2000) and the credit
card fee (0.0246X+1) would bring the total expenses for last year if
the banquet was not included to $7667, or $43 per person.

This is not a joke. Last year, if there was no banquet, SOCG would
have cost $43 per person on average (presumably faculty would be
higher and students lower).

These numbers show that conferences need not be high-budget affairs.
Don't we have some responsibility to the governments that generally
pay for our trips to limit this extravagance? Rock-bottom fees will
certainly help attract students and non-talking faculty to attend
without affecting in any way the exchange of ideas that is the whole
purpose of an academic conference. To socialize we can all just go
drink a beer.

[Ken Clarkson]

I should mutter a belated apology to Tarheels everywhere. If I attribute my outburst to certain traumas experienced in Raleigh at ages nine and ten, it is only to explain, not excuse.