Mea Culpa and Good News

Paul Beame has asked me to post the following message:

I apologize profusely for the consternation that my e-mail to you has caused to you, to members of the SoCG and SIGACT community and to ACM.  I had not discussed our prior plans about co-location of SoCG and STOC with the ACM leadership and this was serious mistake.  It has been clarified that the original concerns from ACM leadership that I had misinterpreted and communicated to you, related only to direct financial support.  I interpreted this far too broadly and then compounded it with a badly worded e-mail to you.  Had I discussed our intended co-location with the ACM leadership, this problem would not have arisen.

As you and I discussed on Friday, SIGACT is very interested in co-location with SoCG and that even if formal co-location were not possible, we were interested in an informal co-location arrangement.  The good news is that contrary to my message, formal co-location of STOC and an independent SoCG is indeed an option and ACM staff will be ready to work with STOC on aspects of such arrangements!  There is a range of options on how this can work and I would like to continue to discuss them with you.

Needless to say, this is fantastic news!  I want to publicly thank Paul and the ACM leadership for quickly clearing up this unfortunate misunderstanding.

And so now, officially: SOCG 2016 will be collocated with STOC in Boston/Cambridge.

Posted in Uncategorized | 4 Comments

Colocation with STOC 2016: Good News and Bad News

Updated August 25: See the bottom of the post.

For almost two years, the steering committees of STOC and SOCG have been considering co-locating the two conferences in 2016, to foster better cross-fertilization between the two communities. The idea received enthusiastic support at the STOC and SOCG business meetings in 2012. After some discussion, the conference leadership settled on Boston/Cambridge as a promising location and convened a join colocation committee to work out the details.

The committee members are Greg Aloupis (SOCG local chair), Venkatesan Guruswami, Sariel Har-Peled, Monique Teillaud, and Daniel Wichs (STOC local chair). Paul Beame (SIGACT chair), Tal Rabin (SIGACT treasurer), and I have also been participating in committee discussions ex officio.

The Boston colocation bid was approved at the STOC 2014 business meeting. The vote at the SOCG business meeting was less enthusiastic, mostly because many of the important logistical details, like precise location and cost, were not yet fixed. (Just like the delay in the SOCG/ACM vote, this lack of detail in the colocation bid is entirely MY fault for not convening the joint committee earlier.) The business meeting vote for SOCG 2016 was tied between Boston and Brisbane, Australia (the only other bid). In fact, there were two votes, both of which were exactly tied. This left the final decision of where to locate SOCG 2016 up to the steering committee.

Since the business meeting, the joint committee has worked out many of the remaining details. The committee agreed on a seven-day schedule, with STOC on the first three days, joint workshops/tutorials held on the middle day, and SOCG held on the last three days. We also discussed possible joint plenary sessions late in STOC and early in SOCG, to be arranged by the respective program committees. Greg (with help from Csaba Toth and Erik Demaine) has located two suitable locations for SOCG and the joint workshops: one at the Stata Center at MIT, the other at Tufts Medical (in downtown Boston, not on the Medford campus where SOCG was held in 2011). STOC will be most likely be held in a hotel in the Back Bay area.

Based on this new information, the SOCG steering committee voted unanimously yesterday to hold SOCG 2016 in Boston/Cambridge, in colocation with STOC.

That’s the good news.

The bad news is the response that I received from Paul Beame, the chair of the SIGACT executive committee, this morning (emphasis added):

This is great, though there is a hitch. We had our scheduled SIGACT executive committee phone conference involving ACM staff yesterday (our first since June). With SoCG leaving ACM we have been told that SIGACT and ACM conferences cannot have any formal arrangements at all with the new conference or do anything official that might support it. (This decision was made at the very top of ACM and not by the staff.) This rules out any joint sessions. . . . It also means that SIGACT will have to end our participation in this formal coordinating group.

Avrim Blum, SIGACT vice-chair, is on sabbatical at UIUC and he can give you more details.

I will spare you my profanity-laden response, but I do want to emphasize the sentence in bold. This is not a decision by the SIGACT executive committee, which has been consistently supportive of colocation even in the face of the SOCG community’s decision to leave ACM. This was not a decision made by Donna Cappo or the other ACM SIG Services staff. This decision was made at the very top of ACM.

It is my personal hope that the two conferences can still come to a reasonable, if unofficial, colocation arrangement, but it is unclear at this point whether this will be possible. Independently of the colocation arrangements, SOCG 2016 will most likely be held in Boston/Cambridge, although even this decision now needs to be reexamined by the steering committee.

Constructive feedback from the SOCG and STOC communities is welcome.

Updated August 25: At Paul’s request, I’ve removed one sentence from his reply, which Paul has clarified was speculation not based on actual ACM policy or instructions.


Posted in Uncategorized | 17 Comments

Suresh’s Response to Vint Cerf

Suresh Venkatasubramanian responds beautifully to Vint Cerf’s request for feedback on how ACM can better serve professional programmers. Read the whole thing!

It’s a little funny though that you’re worried about ACM’s presence in the professional world. Many of us have long assumed that ACM spends most of its focus on that side of the computing community (the excellent revamp of the CACM under Moshe Vardi being the exception that proved the rule). In fact, I’d go as far as to argue that the ACM would be much better served if it were instead to realize how it’s driving itself into irrelevance in a research community that so desperately needs an institutional voice.

My own response is much shorter:

Looking back at my own ACM affiliation, which began in 1967, I was advised by my mentors in graduate school that ACM membership was a mark of a professional and I continue to hold that view. But it seems evident this is opinion not as widespread today. Why not?

I see two big reasons. First, people have become less enamoured with the trappings of professionalism; especially in computing, the view that neckties are a mark of a professional is also not as widespread as it was in 1967. Second, thanks to that little thing you invented called the Internet, we no longer need professional societies to stay connected and well-informed; it’s your fault!

Update Aug 1: See also this response from IT World.


Posted in Uncategorized

Voter Feedback

Participants in the ACM/SOCG vote provided lots of insightful feedback, both in the “other” responses to the second and third question and in the open-ended final question. I’d like to show some of this feedback, on both sides of the main question.

Please don’t be offended if I don’t publish your comments here; this is only a representative sample.  On the other hand, if you recognize your own words and prefer that I not post them publicly, please let me know.

Most of the feedback fell into a few overlapping categories: (1) concern about losing prestige, (2) trust in ACM’s efforts to address our concerns, (3) criticism of ACM, (4) concern about distancing ourselves from the larger community, (5) support for open-access proceedings.

Several people on both sides of the main issue mentioned the potential loss of prestige.

[stay] I work in a small country, where evaluation of candidates for postdoc positions, government-funded projects, etc, is often made by looking at the impact factor of your published papers, how many of your conference papers appear in X database, etc. Even though I cannot agree with these evaluation methods, leaving ACM may penalize researchers from places where evaluations are done in similar ways.
It takes so much effort to have a SoCG paper accepted, that I feel some people cannot afford that it doesn’t count in their CV’s (at least, for some period of time).

[stay] I think we are a small community, and the ACM label (if we like it or not) carries a lot of weight in promotion and tenure decisions. The SoCG community is much smaller than the theory community (STOC, FOCS, SODA) or other applied communities such as the database community (e.g., affiliation-free VLDB). Can we afford leaving now without hurting our young?

[leave] My vote is a “weak” one—I hesitate quite a lot. The opinions of the steering committee and of the past local organizers should probably have more weight in the decision. I agree that there are many arguments in favor of leaving ACM. My only real concern is with respect to the loss of prestige we can expect if we leave ACM. Some countries really view the ACM label as important, and some researchers in our field will be negatively impacted.

[stay] While I do not like the ACM as an organization, and certainly not their policy to increase the cost of the conference significantly, I think that a choice to stay is a more safe choice. It can be that CG as a research field is perceived as a less high quality field 10 years from now by the the general CS research world, just by leaving the ACM. This is not in our control. The quality of the conference itself is in our control, not how it is perceived. I believe the argument that our colleagues in Asia and the US have a better chance of getting tenure if SoCG is with the ACM (I hate that this is the case, but I think it is, so we have to deal with this situation the right way).

A few people advocated trusting ACM, or at least continuing to negotiate with them, now that the community has made its needs clear.  (My own sense is that SIGACT already promised as much as we could reasonably expect to negotiate; concessions from ACM proper were never a realistic possibility.)

[stay] Since ACM officials are now saying that they will take care of the problems, there is no need to leave the organization.

[stay] Given the new leadership change of ACM and their new promises for resolving the past issues regarding SOCG vs. ACM, I think we have made our voices loud and heard, and I believe that they have a good faith to really resolve the past issues/problems. It would be a win-win situation to keep the association with ACM while having the past issues/problems resolved. At this point, I think we should stay with ACM to “harvest” this result.

[leave] We successfully got international SoCGs to work by threatening to leave ACM. We could do the same for the other issues, e.g. by threatening to leave ACM unless they make our proceedings free.

More voters were directly critical of ACM, even while recognizing the prestige issue. Most of this criticism was aimed at the administrative side of ACM; feedback on the researchers who volunteer as SIG officers was generally positive.

[leave] Premier conferences without ACM (or IEEE or SIAM) like VLDB or ICML seem to have no issues without ACM, and enjoy far greater flexibility and autonomy. For historical issues (ACM seems to have provided important early sponsorship and credibility) I would like to stay, but they seem to be no longer useful, and in fact are seeming to become a hindrance. And I don’t see things getting better.

[leave] I find ACM’s claim to the name SoCG appalling, and that alone would make me want to leave them immediately. And try to keep the name.

[leave] I organized SoCG outside the US. I realized that there was no real help from ACM. Contacting them is close to impossible and depending on them implies a big loss of time and effort for a small gain.

[leave] I am torn about leaving ACM; I fear irrelevance if we go it alone, but I’m not sure that ACM promotes relevance. While the SIG volunteers have always been good to us, the staff does not have our interests at heart.

[leave] It is extremely important to clarify that ACM is an organization of scientists, by scientists, for scientists. In the past, support staff at ACM has abused their position to usurp power they are not entitled to, and displayed an appalling lack of responsibility and respect. This cannot be tolerated!

Staff at ACM have to show responsibility for the scientific interests. If we end up staying with ACM (and I’m really sitting on the fence about this), we should
– make clear that it was despite of staff effort, not because
– make clear that things only moved with the help of scientific officers.

Thus, I do not trust staff. My worry is that they will simply sit out this issue, and (ab-)use the next change of office. We need assurances that we can expect continuity, even if the key people change.

[leave] I’m quite on the fence for this one. My biggest argument towards staying is the broad name-recognition of “ACM”, and the community name-recognition of “SoCG”. What propels me towards voting for leaving anyway is that from what I hear and read, ACM and SIGACT have not treated us (and especially the SoCG organizers) well, and get a heck lot of money from us for doing very little for us.

[leave] SOCG is for a truly international community; ACM does not behave like a truly international society.

[leave] I feel the ACM has lost touch with the community, failing to serve its interests (open access, etc.).

Another important issue is the interaction between computational geometry and the broader computer science community. (This is one of the reasons I am strongly in favor of co-locating SOCG with STOC in 2016.  Choosing not to publish our proceedings with ACM does not prevent computational geometers from working more closely with SIGACT and ACM as members.)

[stay] Leaving ACM would make us more isolated from the rest of theory. We are already pretty much isolated…we shouldn’t accelerate it by leaving ACM.

[stay] All of the above PLUS: participation in THE professional society for computer science. Those unhappy with ACM, and there are many reasons to be unhappy with ACM, should work to make our professional society better.

[stay] several of the above, I’d also like to see the CG community closer to the other communities in SIGACT, and leaving ACM doesn’t seem ideal for this.

[stay] ACM is an important, but unfortunately US-centric organization. By staying we can help making ACM more international, if we leave the divide between CS in the USA and the rest of the world only gets larger.

[stay] I’d also like to see the CG community closer to the other communities in SIGACT, and leaving ACM doesn’t seem ideal for this.

Finally, as one would expect from the vote, several people voiced support for freely available proceedings, either officially or by “gaming the system”.

[leave] ACM and IEEE have a growing divide between the academic minority, who provide the papers, and the non-academic majority, who buy them.

[leave] I applaud the attempt to move to LIPIcs. My library does not subscribe to DL so I access SOCG papers by paying for my own individual access… and the move to a truly open access platform is far better. I hope the vote succeeds!

[leave] My main concern is the point highlighted by Pat Morin in his post. The main reason I voted to leave ACM is because I would like everyone to have free and open access to all of our publications.

[stay] Option A: Stay with ACM; put papers on arXiv ensuring “Open-access” with no fee; Play the game that hiring committees do play in many countries, i.e. ACM tag makes your publication worth more.

Option B: Leave ACM; Go “Open-access”; publications “worthless” for bureaucratic committees in many countries.

The obvious choice for me seems to be to stay with ACM while the authors “game the system” by uploading full-drafts on arXiv.

Posted in Uncategorized

A Closer Look at the Voting Results

The recent ACM/SOCG vote asked several additional informational questions in addition to the primary question of whether to stay with ACM or leave.  In addition, PollDaddy reports the country from which each vote was cast.  This additional data reveals some interesting differences between different segments of the community. To preserve anonymity, I will stay silent about groups with less than 10 people (for example: American PhD students, or voters from Singapore).

Update (Jul 20): Please also see Stefan Huber’s plots of this data, by percentages instead of raw counts.


The computational geometry community—and SOCG in particular—has always benefited immensely from the efforts of numerous volunteers.  As expected, a significant fraction of voters indicated willingness to volunteer if the community voted to leave ACM.  More than 30 voters specifically mentioned being willing to help with local organization; the next most common interests were proceedings, record keeping, and publicity. Several people expressed willingness even though they were unclear how they could contribute. Perhaps also as expected, voters who preferred to leave ACM were more likely to volunteer, and vice versa.

Willing to help? Total Stay Leave
Total 185 60 125
Yes 55 18 37
Maybe 92 23 69
No 38 19 19


Five countries contributed more than 10 votes each: Canada, France, Germany, the Netherlands, and the United States. The remaining votes divide roughly evenly into other European countries (Austria, Belgium, Denmark, Greece, Hungary, Italy, Portugal, Slovenia, Spain, Sweden, Switzerland, and the United Kingdom) and everywhere else (Australia, Brazil, China, Hong Kong, India, Iran, Israel, Japan, South Korea, Russia, and Singapore). Yes, I realize the last cluster doesn’t make much sense.

ACM? Willing to help?
Country Total Stay Leave Yes Maybe No
USA 55 29 26 12 22 17
Germany 27 6 21 10 13 1
France 20 1 19 1 13 5
Canada 16 5 11 4 10 2
Netherlands 15 6 9 6 6 1
Other European 39 10 29 10 18 7
Elsewhere 30 11 19 12 10 5

The difference between the United States and other countries is striking.  The US was the only country with more than 10 votes with a majority in favor of staying with ACM.  A few countries with fewer votes also favored staying with ACM, most notably Japan, but most countries outside the US had an overwhelming preference for independence. I think this difference speaks volumes about ACM’s international perception.

French voters were nearly unanimous, presumably thanks to the disastrous experience with ACM in Paris in 2011 (which instigated the vote in the first place).

Recent Attendance

We also asked about recent attendance at SOCG.  As expected, there was a big spread among the voters, and that voters who have attended SOCG more often are more willing to volunteer. More surprisingly, people who have attended SOCG more often were also significantly more likely to prefer staying with ACM.

How many SOCGs ACM? Willing to help?
in last 10 years Total Stay Leave Yes Maybe No
Total 194 65 129 54 92 38
5+ 48 22 26 19 18 9
3–4 47 18 29 12 25 7
1–2 62 15 47 13 31 14
Never 37 10 27 10 18 8

Career Stage

Finally, we asked about academic career stage. Most of the people who answered “other” indicated a position in industry, but we also heard from a few administrators, retirees, and pre-PhD students. Younger members of the community were significantly more likely to prefer independence; tenured faculty accounted for just over half the overall votes but just under two-thirds of the votes to stay with ACM.

ACM? Willing to help?
Position Total Stay Leave Yes Maybe No
Total 193 65 128 54 92 38
Tenured faculty 109 46 63 32 51 20
Untenured faculty 23 3 20 6 13 2
Postdoc 26 5 21 10 10 5
PhD student 21 4 17 0 16 5
Other 14 7 7 6 2 6

Old Farts

Finally, since more than half the voters were tenured faculty, I broke those down further, first by recent attendance, and then for each country with at least ten tenured faculty. Almost everyone who attended 5 or more of the last 10 years has tenure.

The biggest surprise for me is that American tenured faculty preferred staying with ACM by a factor of two to one, exactly the reverse of both the overall vote and the vote among American voters without tenure.  Again, I think this difference reflects the perception of ACM’s importance/prestige: Stronger within the US than globally, and stronger among older researchers than younger.

ACM? Willing to help?
Position Total Stay Leave Yes Maybe No
Tenured, 5+ 42 21 21 17 16 7
Tenured, 3–4 30 13 17 6 16 5
Tenured, 0–2 37 12 25 9 19 8
Tenured, USA 32 21 11 6 13 11
Tenured, Germany 16 4 12 9 6 0
Tenured, France 16 1 15 1 11 4

This difference also highlights the biggest challenge facing the newly independent conference: Maintaining the prestige of our flagship conference, and more generally of the community, without ACM affiliation. This was by far the most popular reason for voting to stay; several people suggested that people may develop a lower opinion of the field simply because it lacks the ACM label.

So much for numerical data.  In the next post, I’ll report some of the insightful narrative feedback that many voters provided.


Posted in Uncategorized | 2 Comments

Results of the Vote

The steering committee for the Symposium on Computational Geometry has concluded its third and final vote on the future relationship between SOCG and ACM. Voting was open for three weeks, from June 24 through July 15, and was made available to all subscribers to the compgeom-announce mailing list.

A total of 202 votes were cast, more than twice as many as in the previous vote.  Here are the results for the main question:

  • I prefer to stay with ACM: 68 votes (34%)
  • I prefer to leave ACM: 134 votes (66%)

Results for the other poll questions are summarized below.

In light of the decisive majority in favor of leaving ACM, the SOCG steering committee (Mark de Berg, David Eppstein, Jeff Erickson, Joseph Mitchell, and Günter Rote), the 2015 local organizers (Bettina Speckman and Marc van Kreveld), and the 2015 program committee co-chairs (Lars Arge and János Pach) have unanimously agreed to immediately move forward with independence, starting with the June 2015 conference in Eindhoven.  SOCG 2014 will be the last iteration of the conference affiliated with ACM.

On behalf of the steering committee, I would like to thank everyone in the community for their input.  I would especially like to thank Paul Beame, Donna Cappo, Wm. Randolph Franklin, Sándor Fekete, Wayne Graves, Pat Morin, and Suresh Venkatasubramanian for their significant contributions to the discussion blog prior to the vote.

The steering committee will publish a more detailed public announcement soon. I also plan to post a more detailed analysis of the voting results later this week.

Now the real work begins!

1. What is your preference for the future organization of SOCG? [202 responses]

  • 134 votes (66%): I prefer to leave ACM and organize SOCG as an independent conference with proceedings published in LIPIcs and with financial backing provided through other means, starting as soon as practically possible.
  • 68 votes (34%): I prefer to stay with ACM, with the understanding that ACM will approve all good-faith requests for in-cooperation status outside the United States. Refusal of in-cooperation status outside the US triggers our immediate departure from ACM.

2. If you voted to stay with ACM, please select your main reason. [72 responses] (A few voters who voted to leave ACM also answered this question.)

  • 46 votes (64%): Prestige of being affiliated with a professional organization
  • 9 votes (13%): Having proceedings in the digital library of a professional organization
  • 8 votes (11%): Other
  • 6 votes (8%): Maintaining continuity
  • 3 votes (4%): Financial backing and insurance

3. If you voted to leave ACM, please select your main reason. [133 responses]

  • 68 votes (51%): Open and free access to the proceedings
  • 23 votes (17%): Reducing costs for the attendees
  • 17 votes (13%): Reducing administrative overhead for the local organizers
  • 14 votes (11%): Other (most common response: “all of the above”)
  • 11 votes (8%): Increased flexibility

4. Independent of your vote in question 1, would you be willing to help out as a volunteer in the event that SOCG does go solo? [185 responses]

  • 92 votes (50%): Maybe
  • 55 votes (30%): Yes
  • 38 votes (21%): No

5. What is your current academic status? [193 responses]

  • 109 votes (56%): tenured faculty member
  • 26 votes (13%): postdoc
  • 23 votes (12%): untenured faculty member
  • 21 votes (11%): PhD student
  • 14 votes (7%): other

6. How many times have you attended SOCG in the last ten years? [194 responses]

  • 62 votes (32%): 1 or 2 times
  • 48 votes (25%): 3 or 4 times
  • 47 votes (24%): At least 5 times
  • 37 votes (19%): Never

Vote distribution by country:

  • 54 votes (27%): United States
  • 27 votes (13%): Germany
  • 20 votes (11%): France
  • 16 votes (8%): Canada
  • 15 votes (7%): Netherlands
  • 6–10 votes each: Austria, Brazil, Japan, Switzerland
  • 1–5 votes each: Australia, Belgium, China, Denmark, Greece, Hong Kong, Hungary, India, Iran, Israel, Italy, Portugal, Russia, Singapore, Slovenia, South Korea, Spain, Sweden, United Kingdom

Posted in Uncategorized | Tagged , | 4 Comments

One week to go. Please vote!

Everyone on the compgeom-announce mailing list should have received an email yesterday with a link to the voting site and a reminder that the voting period ends July 15, one week from today. If you did not receive the reminder (or the initial email), please let me know.

Let me reiterate once again: All members of the computational geometry community are strongly encouraged to vote. It is important that we hear from as many members of the community as possible, especially since one of the concerns with the previous vote was the relatively meager participation.

If you have already voted, thank you! If not, please vote soon.

Posted in Uncategorized