Network of Excellence in Distributed and Dependable Computing Systems
IST Contract No. IST-2000-25088

 
CaberNet Logo

http://www.cordis.lu/ist/home.html
Overview
 What is Cabernet?
 Members
 Staff Directory


Activities
 Latest News
 Plenary Workshops
Radicals Workshops
 Exchanges and Visits
Related Events
 Jobs


Research
 Technical Reports
 Projects
 State of the Art
Vision of RTD
Marketing


Committees
 Links to Industry
 Research & Training
 Information Services
    

7th CaberNet Radicals Workshop

Bertinoro, Forlì, Italy, 13-16 October 2002

Joe Finney, University of Lancaster
Mads Haahr, Trinity College, Dublin
Alberto Montresor, University of Bologna

Aims of the Workshop

The CaberNet Radicals Workshop series aims to foster communication and cooperation between Europe's young distributed systems researchers. The workshop targets researchers who are as yet not fully established in the field of distributed systems. It is aimed in particular at those researchers (most likely post-doctoral) who are in a position to organise inter-node collaborative research projects and subsequent applications for European funding.

The attendance at the workshop has been limited to 26 delegates, just like in previous years, in order to encourage highly interactive discussions. Furthermore, we aimed to achieve a good distribution of attendees from the various CaberNet member nodes, and a reasonable gender balance.

Format of the Workshop

The general format of the workshop was similar to the format proposed by previous workshops. The workshop was composed of a debate session, four presentation sessions and an ongoing poster session.

The debate session was designed to stimulate a discussion about a predefined topic of general interest. The topic was selected according to issues addressed by the delegates, and was centered around the theme "Future Trends for Distributed System Research".

The aim of the poster session was to provide an opportunity for delegates who did not present a position paper to introduce their research interests to their fellow delegates. As opposed to the previous workshop, this incarnation of Radicals featured a special poster session, which was ongoing throughout the entire duration of the workshop, with delegates having highly interactive discussions during coffee breaks between technical sessions. The aim was to foster an increased participation of all delegates, by making the posters a natural topic of many conversations.
Each day was split into several slots, including sessions with technical presentations, the debate and the project sessions. The workshop sessions are discussed in more detail below.

Workshop preparation

The schedule for this edition of the workshop was very tight. The first CFP was released on the 10th of July 2002, with a deadline for submissions targeted for early September 2002. The selection and programme preparation process took two weeks in order to announce the accepted papers and to disseminate necessary information for workshop attendance in the second half of September, just three weeks before the workshop. For personal problems and time constraints, three delegates were not able to arrange their travel plans and arrive on time. They have been substituted by co-authors of accepted papers that were willing to participate in the event.

We received a total of 30 submissions, from which we had to select 26. We applied the criteria of accepting papers written by delegates originating from as many CaberNet member nodes as possible. Nevertheless, we strongly encouraged those young researchers whose papers were not selected, to submit again in future editions of the workshop. From the 26 accepted papers we selected 16 for presentation in technical sessions, and the other 10 were presented as posters. For the selection criteria we took into account the overall quality of the paper, the subject area and, the degree of innovation of the proposed ideas and their potential to foster discussion. The 16 selected papers were divided into four main subject areas, namely "ad-hoc computing", "peer-to-peer systems", and "security" and a more general "distributed system" area.

For statistical records, we can mention that of the 30 submissions we received, four were from authors with completed PhDs (compared to six at the Madeira workshop), and the other 26 from PhD students (or at least candidates to). The delegates were from 18 different CaberNet member nodes and from 9 countries. All of the four rejected papers came from nodes that had also had a paper accepted and were hence represented. Those of the organisers who had attended the Madeira workshop felt that the participants at the Bertinoro workshop were just a little less mature. In addition to two of the organisers, only two of the Bertinoro participants had previously attended a radicals workshop.
Workshop location and facilities

The workshop was held in the University Residential Center of the University of Bologna, located in the small medieval hilltop town of Bertinoro. This town is in Emilia Romagna about 50km east of Bologna. The overall quality of the residential center was very good; the Center is used by the University of Bologna to organize a large number of events, including workshops and conferences, and provides all the facilities needed for organizing a successful event.

Introductory Session

The first session started with the highlight of the aims and objectives of the workshop. The workshop delegates were motivated to create an interactive and lively atmosphere with a greater degree of freedom than experienced in conventional workshops and conferences. This introductory session also included information about funding opportunities and on how CaberNet can fund cross-site visits aimed at further collaboration.

After the introduction, the delegates were asked to present themselves, their work and main research interests. Each delegate was required to prepare a single slide in advance of the workshop with all the relevant information (such as hobbies) in order to present him/herself (and advertise his/her work). It was interesting to see the different types of slides presented, ranging from high-quality design computer-made slides to last minute hand-made ones.

Debate

The Monday morning session of the workshop consisted of an open debate session, being the topic "Future trends of Distributed System Research". Note that the debate was primarily aimed at promoting the development of discussion skills.

The format of the debate was very informal. We started by asking people to identify topics that they felt were important for future distributed systems research. These topics were noted by the organizers on a shared whiteboard and continuously annotated, eventually forming an ad-hoc mind-map, as the debate progressed. Many topics were raised during the debate, but the main two were probably the trend of using biological systems as models for distributed systems and trust in Peer-to-Peer networking.

We felt that while the session eventually gathered some momentum, it took longer than we had expected for many delegates to overcome their initial shyness. We attribute this to the fact that the delegates were just a little less mature than was the case for the Madeira workshop, but also because we had placed the debate fairly early in the workshop programme, namely in the morning on the first day, after the first technical session and before any of the project sessions. For this reason, the delegates had not yet had the experience of working in groups and this seemed to cause them to hold back more than if the debate had been held on the second or third day of the workshop. It should be said, though, that we did find the debate better in terms of interactivity than many debate and panel sessions seen in major workshops and conferences. Overall, we can strongly recommend debate sessions for future workshops, but would advise they be placed later in the workshop programme.

Technical Sessions

The four technical sessions occupied the morning and afternoon of the first day as well as the morning of the second and the third day of the workshop. 20 minutes were devoted to the presentation of each paper. Following the second, third and fourth technical session, the delegates were split randomly into four groups.

Each group was allowed approximately 45 minutes to create a project based on at least two of the papers presented and to prepare a presentation of their project to the other groups. The presentations were required to justify decisions as to the inclusion or exclusion of papers in the projects presented. The session concluded with a vote to discover which of the projects was the most appealing - each delegate voted on their preferred project, excluding his / her own project.

The primary purpose of the technical sessions was to encourage participation and interaction between the delegates. We therefore emphasized that projects were allowed to be radical in nature, sacrificing some practicality for more far-fetched/tenuous concepts and architectures. In this year's edition of the workshop, "projects" with names such as CAR-TO-CAR, SWINY (Share What is Not Yours) and the WHY-Calculus were presented. By and large the projects were quite radical and hence entertaining, generating a comfortable atmosphere and encouraging a modicum of competition between the groups. It was nice to verify that in this workshop many non-native English-speaking delegates were readily available to present the projects and contribute to the good and participative atmosphere. Nevertheless, as usual, and perhaps inevitably, there were some delegates who were more active than others.

Overall, and although the basic structure of these technical sessions was not modified as recommended after the February 2002 workshop, we feel that the objectives of these sessions were fulfilled. It seems that the success of these kinds of sessions strongly depends on the maturity and dynamism of delegates, which fortunately was reasonably good this time.

Poster Session

As in the previous incarnation of radicals, a poster session has been introduced to foster an even better interaction among the delegates. As opposed to the February 2002 workshop, the poster session was a continuous session in the sense that the posters were put on display in the morning of the first day and remained in place for the duration of the workshop. No specific time slot was allocated for the presentation of the posters, but delegates were encouraged to examine the posters. The posters had deliberately been prominently mounted on stands in the middle of the room where the coffee breaks were held. For this reason, they provided a natural starting point for conversations in a relaxed atmosphere and poster presenters ended up spending many coffee breaks discussing their work. We felt this format of the poster session worked extremely well and would not hesitate to recommend it for future workshops.

We did not constrain the format of the posters themselves. However, we encouraged delegates to prepare imaginative designs, to present their ideas in an attractive, simple, objective or even radical way. Most posters indeed followed the suggestion!

The most positive aspect of having posters of non-presented papers was that everyone was able to expose their work and eventually receive some feedback, or at least discuss it with other people during coffee breaks, meals or even during the workshop trip. As opposed to the February 2002 workshop, we decided to include the poster papers in the proceedings.

Evaluation of Success

As already mentioned, this year we had delegates from many CaberNet nodes and from many countries: Greece, Italy, Germany, Holland, UK, Ireland, Portugal, France and Austria.

We believe the workshop achieved its objectives, encouraging a high degree of interaction between the delegates both during the daily formal sessions and the social events (planned and unplanned) in the evening. A questionnaire has been recently disseminated among all participating delegates in order to collect some feedback and/or suggestions about improvements for future editions of the workshop. The results available so far suggest that this was a very successful event with much potential collaboration forthcoming from those who attended. One measure of this is the recent increase in enquiries for CaberNet funding for short visits. It is hoped that these enquiries will translate into applications for funding and indeed, funded trips soon.

Finally, we must mention that in the end of the workshop we voted to elect three delegates as members of the next Radicals Workshop program committee. Depending on the next workshop location, there will be an additional committee member, local to the event place, which will take care of local arrangements. The elected delegates were Rebecca Isaacs, from Microsoft Research, David Akehurst, from the University of Kent, and Lucian Wischik, from the University of Bologna. Congratulations!

Papers

Delegates

1. David Akehurst, University of Kent, d.h.akehurst@ukc.ac.uk
2. Sean Blanchfield, Trinity College, Dublin, Sean.Blanchfield@cs.tcd.ie
3. Malika Boulkenafed, INRIA Rocqencourt , Malika.Boulkenafed@inria.fr
4. Rajiv Chakravorty, University of Cambridge, rajiv.chakravorty@cl.cam.ac.uk
5. Cheun Ngen Chong, University of Twente, chong@cs.utwente.nl
6. Ricardo Corin, University of Twente, corin@cs.utwente.nl
7. Antonio Durante, University of La Sapienza, Roma, durante@dsi.uniroma1.it
8. Giovanna Ferrari, University of Newcastle, giovanna.ferrari@ncl.ac.uk
9. Rebecca Isaacs, Microsoft, risaacs@microsoft.com
10. Arasnath Kimis, University of Cambridge, arasnath.kimis@cl.cam.ac.uk
11. Raimund Kirner, Technische Universitaet at Wien, raimund@vmars.tuwien.ac.at
12. Evangelos Kotsovinos, University of Cambridge, ek247@hermes.cam.ac.uk
13. Yee Wei Law, University of Twente, ywlaw@cs.utwente.nl
14. Pedro Martins, Universidade de Lisboa, pmartins@di.fc.ul.pt
15. Hugo Miranda, University of Lisboa, hmiranda@di.fc.ul.pt
16. Almetwally Mostafa, Université catholique de Louvain, metwally@info.ucl.ac.be
17. Nuno Pereira, Polytechnic Institute of Porto (ISEP-IPP), i960820@dei.isep.ipp.pt
18. Christophe Rippert, INRIA Rhone-Alpes, christophe.rippert@inria.fr
19. Tilman Schäfer, Trinity College, Dublin, Tilman.Schaefer@cs.tcd.ie
20. Jean-Marc Seigneur, Trinity College,Dublin, seigneuj@tcd.ie
21. Paulo Sousa, University of Lisboa, pjsousa@di.fc.ul.pt
22. Vasughi Sundramoorthy, University of Twente, vasughi@cs.utwente.nl
23. Elisa Turrini, Università di Bologna, turrini@cs.unibo.it
24. Arno Wacker, University of Stuttgart, Arno.Wacker@informatik.uni-stuttgart.de
25. Lucian Wischik, Università di Bologna, lu@wischik.com
26. Ge Zhang, University of Kaiserslautern, gezhang@informatik.uni-kl.de

Organising Committee

1. Joe Finney, Lancaster University, finneyj@exchange.lancs.ac.uk
2. Mads Haahr, Trinity College, Dublin, haahrm@cs.tcd.ie
3. Alberto Montresor, Università di Bologna, montresor@cs.unibo.it
4. Jon Warwick, University of Newcastle - Cabernet, jon.warwick@ncl.ac.uk