GLOSS - A Graphical Language Server on the Smalltalk Platform

Abstract

The Graphical Language Server Protocol/Platform (GLSP) has recently emerged as a potential standard for communication between web based graphical modelling clients and model servers. It extends the Language Server Protocol (LSP) developed to support web-based Integrated Development Environments (IDEs) communication with language processing tools (parsers, compilers, linters etc. ). Implementations of GLSP have been shown in Java/Javascript and TypeScript/Node. The authors previously developed a graphical modelling client working with a flexible server capable of handling many model types and realised in Smalltalk (Pharo client, VA Smalltalk server). This utilised a proprietary protocol based upon XML messaging. No GLSP server was available in Smalltalk. The paper describes the design and development of a GLSP server in Smalltalk and the insights gained. GLSP is contrasted with the earlier experience. Advantages of the Smalltalk environment are discussed, as well as suggested enhancements to the GLSP protocol. Future work is suggested.

Published in

IWST 24 - International Workshop on Smalltalk Technologies Lille, France

An Advanced Meta-Meta Model for Visual Language Design and Tooling

Abstract

Visual Languages are widely employed in Enterprise Modelling. Our broader research aims to improve visual language design and facilitate rapid adaptation for specific purposes and audiences with a view to improving Return on Modelling Effort (ROME). Most current repositories and tools have hard coded support for notations and meta model concepts of target languages. A small number of tools have facilities to adapt meta models and notations, but these generally require high technical skills. The model described allows definition of arbitrary meta models and supporting notations in a relatively small meta meta model that can be practically implemented economically using property graph concepts. The model supports advanced concepts, including multi-level modelling. These capabilities facilitate tooling which supports rapid visual language definition, iterative improvement and run-time adaptation for purpose or audience.

Published in

Practice of Enterprise Modelling (PoEM), London, UK