The 19th International Conference on
Principles and Practice of
Constraint Programming
Uppsala, Sweden
September 16-20, 2013

COSpeL: The first Workshop on Domain Specific Languages in Combinatorial Optimization

COSpeL (standing for Combinatorial Optimization SPEcific Languages) is the first workshop on the use of Domain Specific Languages in Combinatorial Optimization, to be held in conjunction with CP 2013 on September 16 in Uppsala, Sweden.


The workshop will take place in the morning. Here is the schedule (and as a iCal file). Each talk is allocated 30 minutes. The articles can be found by clicking on the respective titles. The proceedings are also available as a single file.

Session 1 (9:00-10:30)

9:00 Ozgur Akgun, Alan M. Frisch, Christopher Jefferson and Ian Miguel Automated Modelling and Model Selection in Constraint Programming: Current achievements and Future directions
9:30 Bjorn Regnell and Krzysztof Kuchcinski A Scala Embedded DSL for Combinatorial Optimization in Software Requirements Engineering
10:00 Willem-Jan Van Hoeve Developing Constraint Programming Applications with AIMMS

Session 2 (11:00-12:30)

11:00 Joseph D. Scott Rapid Prototyping of a Structured Domain through Indexical Compilation
11:30 Charles Prud'Homme, Xavier Lorca, Rémi Douence and Narendra Jussien Propagation Engine Prototyping with a DSL
12:00 Guido Tack Embedding MiniZinc

Workshop Description

Domain Specific Languages (DSLs) are programming languages or libraries developed to handle specific tasks. We speak about internal DSL when it is implemented as a library inside an expressive host language. For instance, Ruby, Groovy, Scala, Clojure, Haskell, and Python have been forerunners in encouraging developers to write more succinct yet expressive code. We rather speak about external DSL when it is developed using classical programming language implementation techniques (parsing, compilation, code-generation, ...). The advantages of DSLs are that the concepts of the domain can be expressed more naturally, thus improving and facilitating the user experience.

In the context of combinatorial optimization, DSLs may play an important role to spread knowledge and ease the development of new ideas. We believe that DSLs can serve a big role for the future of optimization, and constraint programming in particular, as they can encode in a simple format the knowledge that is nowadays confined in papers or obscure parts of solvers implementations. DSLs have already been proposed to describe among others models (e.g., OPL, AMPL, AIMMS, Zinc, Essence, Comet, Numberjack), propagation (e.g., Indexicals, grammars and automata), search (e.g., search combinators, non-deterministic search), model transformation (e.g., Conjure, Cadmium)... All those tools have been developed to serve different purposes, but their implementers have probably faced similar challenges in the design and implementation of their language.

The aim of COSpeL is to bring together people interested in the development and use of DSLs in the context of combinatorial optimization. Here are few topics that might be discussed during this workshop:

  • Description of DSL/API for combinatorial optimization (for instance DSL for local search, scheduling, packing, rostering ...).
  • Description/Implementation details of a particular (open source) implementation of a DSL.
  • DSL/API facilitating hybridization (LS + CP, MIP + CP, ...).
  • Standardized DSL/API.
  • Position papers on DSLs:
    • Can solvers become simply interpreters for DSLs?
    • Which domains in CP would need the development of new DSLs?
    • How to design, implement, evaluate, advertise a DSL?
    • What role can play DSLs for the dissemination of CP outside the core community?
    • Does DSL necessarily mean black-box solvers?

Call for paper

We invite submissions related (but not limited) to any of the topics listed above. We welcome both technical and “position” contributions. Taking a broad view on DSL definition, we are also interested into papers discussing APIs. COSpeL is a good place to discuss implementation and design choices not easy to be accepted in a regular conference.

Submissions can be either in the form of extended abstracts (i.e., from 1 to 4 LNCS pages), or in the form of long papers (up to 15 pages). Please submit using our easychair page. A small program committee (see below) will handle the submissions, to ensure overall quality and relevance to the workshop. Each accepted presentation (either short or long) will be allocated the same time slot (25 minutes). Accepted papers and abstracts will also be published in electronic proceedings.

At least one author of any accepted paper must attend the event and the paper will be withdrawn if no such participation is secured with the payment of the workshop dues.workshops/cospel

Important Dates

June 28 July 8: Papers and extended abstracts submission deadline
August 9: Acceptance notification
August 30: Final version due
September 16: COSpeL Workshop


Program Committee

  • Narendra Jussien, Mines de Nantes, France
  • Laurent Michel, University of Connecticut, USA
  • Ian Miguel, University of St Andrews, United Kingdom
  • Jean-Noël Monette, Uppsala Universitet, Sweden
  • Charles Prudhomme, Mines de Nantes, France
  • Pierre Schaus, Université catholique de Louvain, Belgium
  • Guido Tack, Monash University, Australia