Introduction to IWST 2022: a word from the editors
Loïc, Lagadec1,2,3 , Vincent Aranega4,5
1
  Ensta-bretagne, France
2
  Université de Bretagne Occidentale, France
3
  UMR 6285 Lab-STICC
4
  Université de Lille 1, France
5
  INRIA, équipe RMoD


                                         Abstract
                                         The International Smalltalk Technologies Workshop (IWST) is a forum around advances or experience in
                                         Smalltalk, bringing together Smalltalk practitioners since 2009. The IWST aims to stimulate discussion
                                         and exchange of ideas on all aspects of Smalltalk, both theoretical and practical. IWST is a co-located
                                         event with the annual European Smalltalk User Group (ESUG) conference. The 2022 edition of IWST
                                         was held in Novi Sad, Serbia, July 24-26, with Lam Research Corporation as a sponsor.

                                         Keywords
                                         Smalltalk, ESUG, IWST




IWST: International Workshop on Smalltalk Technologies’ 22, August 24–26, 2022, Novi Sad, Serbia
" loic.lagadec@ensta-bretagne.fr (L. Lagadec); vincent.aranega@univ-lille1.fr (V. Aranega)
~ https://www.ensta-bretagne.fr/lagadec/ (L. Lagadec); https://aranega.github.io (V. Aranega)
 0000-0003-3778-3144 (L. Lagadec); 0000-0003-4465-1289 (V. Aranega)
                                       © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
    CEUR
    Workshop
    Proceedings
                  http://ceur-ws.org
                  ISSN 1613-0073
                                       CEUR Workshop Proceedings (CEUR-WS.org)
1. Scope
This report gathers six articles grouped into 3 categories of two papers each. The first one
is dedicated to performance: providing guidelines to achieve high performance, measuring
performance in AI. The second one deals with implementation and debugging in smalltalk
through two innovative axes: model-based delta-programming, and time-travelling debuggers.
The last category is dedicated to refactoring: transformation-based refactoring, and automatic
generation of class comments.


2. Accepted papers
    • How Fast is AI in Pharo? Benchmarking Linear Regression (Cat. 1):
      Like many other modern programming languages, Pharo is extending its applications to
      computationally demanding areas such as machine learning, big data, crypto-currencies,
      etc. This raises a need for fast numerical computation libraries. In this work, the authors
      propose to speed up low-level computations by calling routines from highly optimized
      external libraries, e.g., LAPACK or BLAS, via the Foreign Function Interface (FFI).
    • Design Principles for a High-performance Smalltalk (Cat. 1):
      Smalltalk and related languages have been the subject of many implementations, with very
      different design goals. Most have focused on its state-of-the-art development environment
      and rich class library. In this book, the author focuses instead on performance and/or
      standalone code generation.
    • Towards Object-centric Time-traveling Debuggers (Cat. 2):
      Object-centric debugging aims to make debugging object-oriented programs easier by
      focusing debugging operations on specific objects. Time-shifting debuggers allow devel-
      opers to explore executions in both directions of time. In this work, the authors present
      SeekerOC, a prototype time-shifting debugger that provides object-centric debugging
      support.
    • Using Moose platform for the implementation of a Software Product Line ac-
      cording to model-based Delta-Oriented Programming (Cat. 2):
      Software product line engineering enables the reuse of common functionality to imple-
      ment a set of products. In this work, the authors introduce a framework that supports
      model-based engineering to better manage the variability of product lines at a high level
      of abstraction. An industrial use case is proposed to illustrate the interest of the tools
      whose prototype is presented.
    • Transformation-based Refactorings: a First Analysis (Cat.3.):
      Refactorings are well-known behavior preserving transformations. Little work exists
      on the analysis of their implementation and in particular how traditional refactorings
      might be composed from smaller, reusable, parts.In this work, authors focus on the
      possibilities to reuse transformations independently from the behavior preserving aspect
      of refactoring.
    • Can we automatically generate class comments in pharo? (Cat. 3, Best Paper):
      Code comments support developers in understanding and maintaining codebases. In the
      Pharo environment, code comments are the primary form of code documentation and
      typically convey information ranging from high-level design descriptions to low-level
      implementation details. In this work, the authors adopt the stereotype-based approach to
      automatically generate class comments in the pharo programming environment.


Acknowledgments
We would like to thank the Lam Research Corporation for their financial contribution to provide
awards for the best papers.