A new record

Now, that’s it. The program for the upcoming Eclipse Summit Europe 2010 is final. In a few days, maybe even only hours you will be able to see the conference schedule. More than 240 e-mails have been sent after Anne pushed the “Big Button” this afternoon letting people know if their submissions where accepted or not. 242 submissions is a new all time record for ESE after 206 proposals from last year. I think this is great! And I want to thank everybody who submitted something, especially those people whose talks were now rejected. I know that many people are now unhappy about their talk being not in the program. In fact, I already got some messages indicating that only minutes after the mails where send out. Obviously, people are curious to know why their session has been rejected and asked for more transparency. So I will try to explain a bit what has been going on and how we (this is Boris Bokowski, Christine Mitterbauer, Doug Clarke, Ed Merks Peter Siwon and myself) came up with the program which is now online.

In most of the cases the talks were not rejected because of their quality but because of the lack of rooms and time.  ESE is a three-day low-cost conference. In the location we have 5 rooms which can hold 40 to 730 people. The first day is reserved for tutorials and symposia. Each tutorial and symposia is scheduled to be 4 hours. This year we can offer interesting 5 tutorials in the morning and 4 great symposia in the afternoon. Tuesday and Wednesday are then classical sessions where one room is reserved for sponsored talks. Sponsored talks are as well technical sessions, sometimes with, sometimes without advertisement. They are necessary to keep the fees low.

On Wednesday we start at 9:00 in the morning; the last session closes at 6:30 followed by a poster reception and BoFs which you can propose on site. On Thursday we again start at 9:00 and close at 16:00 as many attendees need to catch their trains or plains to get back home again. This gives us a bit more than 11 hour of technical content for those two days all together (given that we need some breaks and food in between as well J). In order to keep the conference fees minimal we can only give away a very limited amount of free passes. We as the program committee decided that we want to have as many sessions possible, but still allow the speakers to have enough time to deliver insightful presentations. This is why we decided to have 8 long talks (50 min each) 52 short talks (lasting 25 min) and no lightning talks. This decision had to be made even before the submission system was opened. Now you know the cornerstones on which our decisions are based on.

As in recent years we had 5 categories – E4, Embedded, Modeling, Runtime and Other / New & Noteworthy. Although we had a fairly large spreading on the number of submission per category (Other/NN had the most, followed by Modeling) we wanted to have roughly the same amount of talks per category. Later, as the program continued to shape this changed a bit, but in general I think this idea is still very much reflected in the program. In total we had to decide for 60 talks out 240 submissions which means that we could only accept 1 out of 4! The criteria to accept a talk was to attract as many people as possible. This means that we tried to figure out what is currently of high interest in the community, which projects have been very active during the last year and thus can tell a story, which speakers are known to deliver great presentations, where are new projects with high potential which are currently only known to few people, … But in addition to that we wanted to have some real user stories in. We think that this as well adds high value if people know not only *how* things can be done but also get some inspirations on *what* can be done with Eclipse technologies. And maybe even the committers (so the people typically giving the presentations) can have some take-away from that and know where they or their projects can improve. I am grateful for having Boris, Christine, Doug, and Ed with me on the PC this year. Everybody of them is deeply involved in the Eclipse community and has deep insights into the projects and the very diverse Eclipse ecosystem. Together we managed in about 20 conference calls and many, many hours of work to read carefully through all submissions, to bring them into an order, to discuss them where we had different opinions, to ask for clarification where necessary and finally, to decide on what we think will be the best program for this year. Being a member of the program committee is for sure an honor but it is as well a lot of work, so once again, kudos to all PC members and thank you for the time you invested. At this point I’d also like to thank Anne for helping us with the various organizational issues we had to discuss and deal with.

I hope you got some insights on how we worked, how we made our decisions and maybe as well why we took them that way. With the many great submissions I think we could have easily filled another 2 days. If your talk was rejected this time don’t give up and give it a try next time again. Or maybe you want to do a poster presentation? The submission system for that is still open. Your next chance for a talk is EclipseCon 2011. The more proposals a conference gets, the harder the job of the program committee is but as well, the better the program will end up.

By the way, have you noticed that we will have 3 exiting keynotes this year? I am really looking forward to every single one of them. And not only to them :-) Hope to see you all at ESE again to have some fun!

6 Responses to “A new record”

  1. 1 Chris Aniszczyk

    Thanks for organizing everything Bernd, I know how hard it is given the start of getting things ready for EclipseCon 2011!

    I’m really looking forward to ESE this year!

  2. 2 Konstantin Komissarchik

    Could you touch on how the tracks were selected? I find the lack of Tools track troubling. I am sure folks from JDT, PDE, WTP, Mylyn, etc. have stories to tell too.

    The trouble with tracks in general is it is inevitably going to cause friction. Why is one area more important than another to have a track and therefore higher likelihood of accepted talks?

    My other comment is on format. Given how small the conference is in terms of number of days times number of rooms, and given the high number of submissions, I wonder if it would have been better to do away with tutorials and maybe even long talks. Each tutorial “costs” eight short talks. Each long talk costs two short talks. If my math is correct, going 100% on short 25-min talks would have added 38 sessions to the program. That’s a lot of additional space to give more corners of the ecosystem time to tell their stories without raising the costs.

  3. 3 Oisin Hurley

    Bernd, congratulations to you and the team for constructing the conference program. It’s a difficult task, in which you can’t avoid annoying some (maybe many!) people.

    I partially agree with Konstantin – I’d personally prefer to provide only 25 minute talks (so no long ones), but I would keep the tutorials. You know my thoughts on that already ;-)

    The tracks part is an interesting one. Tracks are used not just as a container for organizational purposes, but have a promotional element as well. The problem with a “Tools” track is that it is too generically specified – but I get your point, Konstantin, that in the selection of tracks an attendee might be puzzled where to find the latest WTP innovations. One thing that would help attendees would be blog entries by known luminaries giving guidance on what sessions they will attend.

  4. 4 Bernd

    Hi Konstantin,
    Thank you for your comments. Let me try to answer your questions.
    You are right. We could have added a tools track. However the idea of the tracks is more to get a rough overview rather than having a one-to-one relationship to Eclipse Top Level Projects. In order to avoid the friction you are talking about, we made sure that every talk was reviewed by at least two PC members. By that we tried to not reject a talk just because one person did not like it or didn’t find it interesting enough. Just as an example, we decided that WTP belongs to Runtime (although it is not an Eclipse RT project but is certainly RT related. Same holds true for PDE and JDT). The reason for not having a specific tools category was that we had bad experience with that in recent years. The problem was that in the end everything is a tool. So people used it as a generic category and not as something which is related to Eclipse Tools (even though the description said something different).
    To be entirely honest with you, I am not a big fan of the categories at all. I would prefer some kind of free tagging but this is not (yet) supported in the submission system.
    The second reason to have only few categories (not saying that one more would have been a problem) is to give some guidance to the attendees. The more categories we have, the harder it is for the ones which are new to or not so familiar with the internal structure of Eclipse to follow.

    So to make a long story short: If you submitted a talk in a wrong category we moved it and still considered it.
    So now to your comment on the format: We thought about getting rid of symposia and tutorials initially but decided not to. The reason is that those formats are very popular with attendees. The second constrained we had was the amount of free passes we were allowed to spend. The Eclipse Foundation wants ESE to be a cheap, affordable conference. (And I think this is great btw.) This however restricts the whole budget. Initially we had even more long talks and even less short talks – they were split equally as far as I remember. We could only change that because of some companies and individuals were willing to do presentations without the guaranteed discounts. Please keep in mind we had to come up with that calculation before the first talk was submitted. Now let me be clear here: We did not prefer those companies or talks. We came up with a first version of the program. After that we asked some of the speakers or their company representatives if they would be willing to give the talk even if they would not get the discount. Some agreed, some were keen on getting the discount. With that we could go ahead and split most of our long talks into two short talks each. On the other hand, there were still some submissions were we as the PC thought that they deserve a long slot. This is why we did not want to get rid entirely of those anyway. The way we have it not I think is a good balance.

    Best regards,

  5. 5 Bernd

    Thanks Oisin,

    I agree with you. I also prefer the 25 minute talks. But I think some talks are worth the 50 minutes. As already stated in the reply to Konstantin, we had as well some budget constraints which initially forced us to have even more long talks. With the 8 long talks now, I hope we found a good balance :-)

    For the track part, I as well agree with you, but that is what I already replied to Konstantin :-)


  6. 6 Lynn Gayowski

    Excellent job Bernd! Congratulations to you and the rest of the program committee – I know it’s a tough job.

Leave a Reply