Boston, October 22nd, 2018
It is my great pleasure to welcome you to Boston for SPLASH 2018. SPLASH is a joint event, sponsored by the ACM and SIGPLAN, that combines four conferences (OOPSLA, SLE, GPCE, and Onward!), one symposium (DLS), fourteen workshops and two special tracks (SPLASH-I and SPLASH-E). SPLASH runs every year, alternating between North America and Europe. With close to six hundred expected attendees, SPLASH is the largest academic gathering of researchers, practitioners, and students interested in programming technologies.
In addition to traditional SPLASH events such as the SPLASH poster track, the Student Research Competition, the SPLASH-E educational track, the Doctoral Symposium, the SPLASH-I applied track, SPLASH’18 has a few innovations. In cooperation with the organizers of the Programming Language Mentoring Workshop, there are three Mentoring Breakfasts. These breakfasts allow students to meet established researchers and to discuss topics ranging from how to start doing research to questions of work-life balance. The SPLASH-I talks are more numerous and have been extended to one hours slots. Moreover, SPLASH-I is organized jointly with six Meetup group sessions that are held in the evening at the MIT and Northeastern University. To encourage a diversity of backgrounds, developers were offered a number of complimentary tickets to attend SPLASH-I. The conference banquet has been democratized and extended; this year a social hour will be held at the end of each day, and all attendees will be invited to participate. On the last day, a t-shirt social will mix drinks with a raffle for the remaining student volunteer tees.
Preparing SPLASH took the better part of a year with thirty volunteers and an additional fifty students on site. Without their help, enthusiasm, and dedication, SPLASH wouldn’t happen. I am particularly grateful to Konrad Seik and Paley Li for chairing the poster session, Ben Greenman for handling housing and restaurant related issues, Tucker Taft and Heather Miller for chasing after sponsors, Aviral Goel and Filip Krikava for caring after the web site, Benjamin Chung and Leif Andersen for leading video recording, Jonathan Bell and Celeste Hollenbeck for publicizing and agitating on social media, Arjun Guha and Alex Potanin for selecting and looking after the workshops, Karim Ali and Michael Carbin for selecting SPLASH-I speakers and meetup groups, Juliana Franco and Petr Maj for leading the student volunteers, Tijs van der Strom and Fabio Niephaus for overseeing the proceedings, Justin Slepak for worrying about accessibility, Shan Shan Huang and Jay McCarthy for running the student research competition, and Philipp Haller and Benjamin Lerner for, respectively, chair the doctoral symposium and the SPLASH-E track. In addition, we benefited from Eddie Koehler’s HotCRP for submissions, the web site was built with the Researchr system created by Eelco Visser and Elmer van Chastelet, registrations were handled by Carol Mann and Mike Moshell of Registration System Labs, proceedings were prepared by Dirk Beyer’s Conference Publishing Consulting, MeetEcho recorded the OOPSLA talks and keynotes, our logo and t-shirts were designed by Kristýna Mothejzíková, and Annabel Satin managed it all. At ACM, Abigail Sumandal an John Otero provided logistical support for contracting and dealt with the vagaries of the venue. At SIGPLAN, Micheal Hicks and Jens Palsberg provided moral and financial support.
The rest of this note reports on the conference from an organizer’s perspective.
SPLASH’18 has 663 registered attendees from 36 countries. The following graph shows registrations over time. Industrial researchers and software developers account for 165 registrations. Postdoctoral researchers and faculty members are 228, and students for 259. The October 1st early registration marks the clear inflection point in the growth of registrations – no surprise there.
While SPLASH is international, most attendees are US residents (78.8%). The European countries with a larger representation are Germany, the United Kingdom, the Netherlands, Belgium, Switzerland and France. The countries with 5 or fewer attendees are: Australia, Austria, Brazil, Cambodia, Chile, China, Denmark, Georgia, Greece, Hong Kong, India, Italy, Luxembourg, Malta, New Zealand, Norway, Peru, Portugal, Slovenia, South Africa, Spain, Turkey, United Arab Emirates.
| Country | n |
|---|---|
| USA | 407 |
| Germany | 43 |
| Canada | 38 |
| United Kingdom | 23 |
| Netherlands | 19 |
| Japan | 14 |
| Belgium | 13 |
| France | 13 |
| Switzerland | 13 |
| Sweden | 10 |
| South Korea | 8 |
| Israel | 7 |
| Czech Republic | 6 |
Within the US, most of our attendees are from New England. California and New York are also well represented. States with fewer than 5 registrants include AL, AZ, CO, CT, DC, FL, GA, IA, KS, ME, MI, MN, MO, MS, NC, NE, NH, NJ, NV, OH, OR, PR, TN, TX, VT. These numbers suggest that it is beneficial to locate SPLASH in a state with a strong PL community.
| State | n |
|---|---|
| MA | 187 |
| CA | 48 |
| NY | 35 |
| IL | 17 |
| PA | 16 |
| WA | 15 |
| VA | 12 |
| IN | 10 |
| MD | 7 |
| RI | 6 |
| UT | 5 |
Over half of SPLASH attendees (55.9%) are first time attendees – a combination of PLMW students (33 out of 303), new graduate students (127), as well as newcomers from academia and industry (respectively 68 and 75). The first-timers are 35.4% Americans, and 46.9% of those are local to Massachusetts. Close to a quarter of attendees have been at 3 or more SPLASH conferences.
Gender of attendees is reported as Male, Female and Other. We focus on ‘non-male’ attendees; missing data is inferred heuristically when possible. The following graph shows that 16.8% of SPLASHers identify as non-Male. This is slightly higher than previously reported numbers.
To get more insight, we breakdown attendees further. The Programming Language Mentoring Workshop (PLMW) has the highest proportion of non-male attendees, 46.1% are non-male. This can be explained by the fact that diversity is emphasized in advertising and selection. Next, graduate students are 20.4% non-male, a number higher than in many Computer Science departments in the US. There is a significant drop-off in diversity with academics; faculty members and postdoctoral researchers are only 13.6% non-male. Industry is even lower with 6.2% of non-males. It is interesting, but not surprising, that diversity decreases with repeat attendance - repeat attendees tend to be professors and professionals.
The following table show a per-event breakdown of registration. The numbers indicate how many attendees expressed interest in a particular event, this not necessarily an accurate picture of attendance.
| event | regs |
|---|---|
| PLMW | 52 |
| VMIL | 50 |
| GPCE | 47 |
| BLOCKS | 45 |
| SLE | 45 |
| DLS | 44 |
| LIVE | 44 |
| REBLS | 34 |
| HILT | 30 |
| PLATEAU | 30 |
| AGERE | 27 |
| META | 23 |
| TURBO | 17 |
| OCAP | 17 |
| DSLDI | 16 |
| NJR | 12 |
| SLEBOK | 10 |
| GRACE | 8 |
The first challenge in running SPLASH is to pick a venue. That choice drives many other decisions. In 2018, to keep costs down, our original choice was Northeastern University, but that was voted down by the Steering Committee. Instead, we settled on the Boston Park Plaza as it has sufficient space, is available in the late Fall, and offered prices that would not bankrupt SIGPLAN. With a venue, other choices follow. The hotel requires a minimum Food & Beverage (F&B) expenditure and a guarantee of 760 nights. These requirements encouraged us to keep social events on site and steer attendees towards staying at the Plaza. After reviewing the costs of banquet venues, we decide to go without and instead have social hours (on site of course). Rooms are set at $239 guaranteed rate, but the hotel surprised us by limiting the number of double rooms they let our attendees book.
The registration fees are kept at the 2017 level, so the full week costs $575 for students and $1,050 for members of the ACM. From our attendees, 40 are funded to attend the Mentoring Workshop; their registration, travel, and accommodations are fully covered by sponsors. Furthermore, 52 have free registrations thanks to the Student Volunteer program. Additional support is provided by the SIGPLAN PAC program. No other free registrations are offered; all chairs pay their way.
Registration revenues over time show that most revenues come in by the early registration deadline, October 1st, and revenues tapper off after that.
Even with those incomes, the pre-event budget, shown here in a shortened form two weeks before the conference, indicates SPLASH will lose money. The organizers worked hard to keep costs down, but constraints such as F&B minimums and high audio and video costs are hard to offset. As an example, the traditional “ice cream social” will not feature ice cream as the hotel charges $23 for one ice cream cone.
| Income | Expenditures | ||
|---|---|---|---|
| Registration | $250,000 | Food & beverages | $146,114 |
| Sponsorship | $51,985 | Audio & video | $58,445 |
| Open access fees | $6,000 | ACM allocation | $44,780 |
| Miscellaneous | $5,905 | Room penalty | $21,271 |
| Proceedings production | $13,000 | ||
| Credit card fees | $10,249 | ||
| Program committee meeting | $9,010 | ||
| Speakers travel | $8,780 | ||
| Onsite staff | $8,391 | ||
| Total | $313,890 | Total | $320,040 |
The above costs do not account for the substantial infusion of cash from SIGPLAN: event management ($10K), registration ($12K), video recording ($10K), mentoring workshop ($15K), gold open access, and PAC travel grants ($20K). Some of the funds provided by SIGPLAN come from the digital library revenues, but, as open access is becoming the norm these revenues will dry up. Thus to sustain itself SPLASH will either have to raise costs or cut services.
Data is available for six years of SPLASH, 2013 (Indianapolis), 2014 (Portland), 2015 (Pittsburgh), 2016 (Amsterdam), 2017 (Vancouver) and 2018 (Boston), paints a picture of the evolution of the conference. The attendance over these years ranged 482 and 608 attendees. Overall the attendance has been trending down, especially so when compared to the early 2000s. Other events compete with SPLASH for the attention of practitioners, as for academics their participation is a function of the size of the community and the number of accepted papers (at the main conference and at workshops). Location also plays a role, but one slightly harder to predict – comparing Portland and Indianapolis it is unclear why the Midwest proved more appealing.
One encouraging observation is that diversity has improved over the years. The following graph compares the proportion of men attending SPLASH over the years. The chart shows a clear, and hopefully accelerating, trend towards more diversity. A difference of ~9% in six years is significant, there were 58 women in 2013 and there are 92 this year.
That trend is most pronounced by students as shown in the next graph. For this comparison we include PLMW attendees. The proportion of male student decreased from 88% to 72% in six years. While remarkable, is this actually representative of the student population in the field? It is conceivable that these numbers are skewed by the selection process of PLMW and of student volunteer, in both cases diversity is favored.
A less biased indicator is the ratio of non-males in postdoctoral researcher and faculty members attending the conference. This number is not affected by the conference organizers, with the exception of invited speakers. Here the is just as clear, we went from 8% non-male to 16% in six years.
SPLASH has always been an exciting place for me, always full of new ideas and interesting people, I hope you find it a welcoming community and come back many times.
Jan Vitek
Comfy/General Chair
Professor Northeastern University