Published: 2nd October 2021
DOI: 10.4204/EPTCS.347
ISSN: 2075-2180


Proceedings 14th
Interaction and Concurrency Experience
Online, 18th June 2021

Edited by: Julien Lange, Anastasia Mavridou, Larisa Safina and Alceste Scalas

Invited talk: Protocols, Timeouts, and Failures
Laura Bocchi
Invited talk: Execution and Planning of Distributed Systems Reconfigurations - Concurrency and Formal Aspects
Helene Coullon
Processes, Systems & Tests: Defining Contextual Equivalences
Clément Aubert and Daniele Varacca
Towards Generalised Half-Duplex Systems
Cinzia Di Giusto, Loïc Germerie Guizouarn and Etienne Lozes
Deadlock Freedom for Asynchronous and Cyclic Process Networks
Bas van den Heuvel and Jorge A. Pérez
Decomposing Monolithic Processes in a Process Algebra with Multi-actions
Maurice Laveaux and Tim A.C. Willemse
A Semantic Model for Interacting Cyber-Physical Systems
Benjamin Lion, Farhad Arbab and Carolyn Talcott


This volume contains the proceedings of ICE'21, the 14th Interaction and Concurrency Experience, which was held online (due to the COVID-19 pandemic) as a satellite event of DisCoTec'21. The previous editions of ICE were affiliated with ICALP'08 (Reykjavik, Iceland), CONCUR'09 (Bologna, Italy), DisCoTec'10 (Amsterdam, The Netherlands), DisCoTec'11 (Reykjavik, Iceland), DisCoTec'12 (Stockholm, Sweden), DisCoTec'13 (Florence, Italy), DisCoTec'14 (Berlin, Germany), DisCoTec'15 (Grenoble, France), DisCoTec'16 (Heraklion, Greece), DisCoTec'17 (Neuchâtel, Switzerland), DisCoTec'18 (Madrid, Spain), DisCoTec'19 (Lyngby, Denmark), and DisCoTec'20 (virtual).

The ICE workshop series features a distinguishing review and selection procedure: PC members are encouraged to interact, anonymously, with authors. This year these interactions took place on the OpenReview platform which combines paper selection features with forum-like interactions. As in the past editions, the forum discussion during the review and selection phase of ICE'21 considerably improved the accuracy of the feedback from the reviewers and the quality of accepted papers, and offered the basis for lively discussion during the workshop. The time and effort spent on the interaction between reviewers and authors is rewarding for all parties. The discussions on OpenReview make it possible to resolve misunderstandings at an early stage of the review process, to discover and correct mistakes in key definitions, and to improve examples and presentation.

The 2021 edition of ICE included double blind reviewing of original research papers, in order to increase fairness and avoid bias in reviewing. Research papers were blinded for submission, with authors' names and identifying details removed, and authors were given anonymous access to a dedicated section of OpenReview. Each paper was reviewed by three PC members, and altogether 5 papers were accepted for publication − plus 4 oral presentations which are not part of this volume. We were proud to host 2 invited talks, by Helene Coullon and Laura Bocchi. The abstracts of these talks are included in this volume, together with the final versions of the research papers, which take into account the discussion at the workshop.

We would like to thank the authors of all the submitted papers for their interest in the workshop. We thank Helene Coullon and Laura Bocchi for accepting our invitations to present their recent work. We are grateful for the efforts of the PC members:

We thank the ICE steering committee and the DisCoTec'21 organizers, in particular the general and workshop chairs, for providing an excellent environment for the preparation and staging of the event: a remarkable feat, especially considering the challenging circumstances due to the COVID-19 pandemic. Finally, we thank the editors of EPTCS for the publication of these post-proceedings.

Protocols, Timeouts, and Failures

Laura Bocchi (University of Kent, UK)

In the last few years, several efforts were directed at establishing relationships between models for real-time asynchronous interactions and their implementations so to reflect correctness of time-sensitive models into their implementations. I will discuss contributions and limitations, focussing on the way unpredictability (e.g., jitter, failures) is accounted for. I will discuss ongoing directions, in the context of the STARDUST project (Session Types for Reliable Distributed Systems) towards detecting grey failures, as a first step towards better recovering strategies for actor-based systems.

Execution and Planning of Distributed Systems Reconfigurations - Concurrency and Formal Aspects

Helene Coullon (IMT Atlantique, FR)

Large distributed software systems are nowadays built in a component-based approach (e.g., service-oriented architectures or microservices) that offers a convenient way to structure large applications. Indeed, isolating functionalities in components and building systems through composition greatly enhances the adaptability and scalability of applications, two important requirements for many organizations. This approach is also promoted by the massive adoption of highly distributed computing infrastructures such as cloud and edge computing. However, the advantages of distributed architectures come at the price of increased complexity and an array of technical challenges related to observability, coordination, maintenance, etc. For many years deployment and management procedures were written in an ad-hoc fashion through "readme" files and bash scripts and very was often triggered manually by an operator. But, in recent years the DevOps community (both from academia and industry) has significantly improved this area by bringing programming support and software engineering properties to distributed software management.

In this talk, I will explore, through three main contributions, the concept of reconfiguration of component-based distributed systems, in particular, the planning and execution of such reconfiguration. These topics will be studied from two perspectives: the concurrency and efficiency of reconfiguration; and formal and safety aspects of reconfiguration. To this end, I will first present the formal reconfiguration model Concerto in which: (1) the lifecycle of each component is programmable with concurrency, and (2) reconfigurations are constructed by composing component lifecycles and submitting asynchronous behavior requests to components. Next, I will present a contribution that uses timed Petri nets and model checking to ensure qualitative and quantitative properties on a subset of Concerto. Finally, I will present an ongoing work on the automatic synthesis, by using SMT (Satisfiability Modulo Theories), of Concerto reconfiguration procedures from a partial target configuration.